(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Assessing patterns and processes of landscape change in Okefenokee Swamp, Georgia"

ASSESSING PATTERNS AND PROCESSES OF LANDSCAPE CHANGE 
IN OKEFENOKEE SWAMP, GEORGIA 



By 

CYNTHIA SMITH LOFTIN 



-4 



A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL 

OF THE UNIVERSITY OF FLORIDA IN PAR-HAL FULFILLMENT 

OF THE REQUIREMENTS FOR THE DEGREE OF 

DOCTOR OF PHILOSOPHY 

UNIVERSITY OF FLORIDA 

1998 



Copyright 1998 

by 

Cynthia Smith Loftin 



To my family: your love, patience, and encouragement made this work possible. 
In memory of my friend, Millicent Quammen, who was an inspiration. 






ACKNOWLEDGMENTS 

I am indebted to many individuals who contributed to this work. My academic 
advisory committee provided direction in my graduate studies, which has contributed to 
my continuing development as a scientist. The encouragement, tolerance, interest, and 
counsel of my advisor, Dr. Wiley Kitchens, permitted me to explore thoroughly my 
research topic, which contributed significantly to its successful conclusion. I am grateful 
for his continued confidence. Dr. Ramon Littell assisted with design and analyses of the 
vegetation species-hydroperiod and seed bank studies; I appreciate his patience, time, 
and assistance. Dr. Jack Putz's keen, editorial eye forced me to clarify many aspects of 
this document; his thoroughness and frankness are appreciated. Dr. George Tanner and 
Dr. Loukas Arvanitis contributed perspectives and suggestions that improved this 
dissertation. I want to thank them for their interest and time. Dr. Franklin Percival 
attended my defense and provided thoughtful and humorous discussion throughout my 
doctoral program; I appreciate his time and encouragement. 

Many students, technicians, and friends assisted with field work, compilation of 
GIS and spreadsheet databases, and GIS model development and analyses. My sincere 
appreciation is extended to C. Depkin, J. Brookshire, P. Owen, D. Evenson, D. O'Neill, 
K. Williges, N. Ansay, P. Gonzalez, J. Aufmuth, J. Kitchens, J. Halblieb, and J. Loftin for 
their tireless dedication. B. Raspberry of the USFWS assisted with the GPS topography 

iv 



survey; without his involvement, the hydrology study would not have been possible. C. 
Trowell made his summaries of the Okefenokee Swamp human history readily available; 
I want to thank him for bringing many details to my attention. Fellow graduate students 
and coop unit employees L. Brandt, J. Silveira, C. Allen, K. Rice, A. Garmestani, W. 
Bryant, D. Hughes, and B. Fesler contributed thoughtful discussion, administrative 
assistance, and much-needed levity. L. Pearlstine guided me through my initial foray 
into the world of GIS, and W. Hyde provided excellent computer systems management; I 
am grateful for their patience. 

The staff of the Okefenokee National Wildlife Refuge deserve my deepest 
gratitude for their assistance. They permitted unlimited access to records and the swamp, 
as well as assisted in GPS surveys, water level recorder maintenance, and provided 
housing and equipment upon request. S. Reeves, S. Aicher, L. Mallard, J. Burkhart, S. 
Davis, R. Phernetton, T. Hiding, S. Jones, and C. Thompson, as well as many others, are 
dedicated stewards of that natural treasure. Foresight of the USFWS and refuge 
managers to recognize that a study of the effects of the Suwannee River sill on the 
swamp hydrology and ecology was needed and to provide sufficient funding to address 
the salient questions was integral to initiation of this study. 

My family continues to give encouragement and love, which have fueled me 
throughout this effort; I want to thank Dorothy for her patient, loving care of Steven, and 
Kelly for her friendship and super babysitting. Finally, I thank Jim for his understanding, 
interest, and love, which enabled me to attack this project, and without whom I could not 
have completed it, and Steven, who has shown me another facet of life. 






TABLE OF CONTENTS 

ACKNOWLEDGMENTS iv 

LIST OF TABLES x 

LIST OF FIGURES xx 

ABSTRACT xxviii 

CHAPTERS 

1 THE OKEFENOKEE SWAMP AND THE SUWANNEE RIVER SILL 1 

Ecosystem Hierarchies and Scales 5 

Driving Functions 11 

Disturbances, Heterogeneity, and Succession 15 

Monitoring Landscape Change 19 

The Okefenokee Swamp Ecosystem 20 

Human Modification of Okefenokee Swamp 28 

2 DATABASE ORIGIN AND DEVELOPMENT 34 

Data Sources and Extent 34 

Swamp Water Level Data 36 

Water Level Recorder Performance 39 

Estimation of Missing Water Level Data 71 

Precipitation Gauge Network Assessment 83 

Background 83 

Methods 87 

Results of Precipitation Network Assessment 89 

Discussion of Precipitation Network Analysis 106 

Estimation of Missing Precipitation Data 107 

Estimation of Evapotranspiration, Inflow, and Outflow Data 108 



VI 



Swamp Basin Delineation and Characterization 120 

Effects of the Suwannee River Sill on Swamp Water Level Conditions 130 

Topography Surface 146 

Collection of Point Elevation Data 150 

Surface Interpolation 160 

Topography Surface Description and Trends 164 

Satellite Imagery Classification and Accuracy Assessment 1 72 

Image Preparation 173 

Image Classification 176 

Image Classification Accuracy Assessment 178 

Image Classification and Accuracy Results 193 

Interpreting the Accuracy Assessment 199 

Applying Image Classification Procedures 202 

OKEFENOKEE SWAMP HYDROLOGY MODEL 207 

Introduction 207 

Methods 209 

Model Objective 209 

Model Overview 210 

Model Data Sources 220 

Precipitation 220 

Evapotranspiration 221 

Creek Inflow Volumes 221 

River Outflow Volumes 222 

Water Depth and Topographic Surfaces 223 

Data Surfaces Used for Model Assessment 223 

Model Manipulation and Assessment 225 

Wildfires in the Area Affected by the Suwannee River Sill 227 

Vegetation in the Area Affected by the Suwannee River Sill 228 

Results 229 

Area Affected by the Suwannee River Sill 229 

Model Accuracy: 1980-1993 229 

Model Responses to Sill Manipulations 267 

Regional Hydrologic Trends 314 

System Sensitivities 348 

Wildfire Occurrence 361 

Vegetation Change 364 

Discussion 375 

Model Performance 375 

Effects of the Suwannee River Sill 379 

System Sensitivities 382 



vn 



4 LANDSCAPE LEVEL VEGETATION CHANGES IN OKEFENOKEE 

SWAMP 385 

Introduction 386 

Dynamics of Swamp Vegetation 386 

Sill Affected Vegetation Change 390 

Methods 391 

Logging Tramlines 391 

Pre-Logging Vegetation (1850-1890) 395 

Post-Logging Vegetation (1952) 398 

17 Years With-Sill and 22 Years Post-Fire Vegetation (1977) 410 

30 Years With-Sill and 35 Years Post-Fire (1990) 411 

Wildfire Burn Area Maps 411 

Map Comparisons 415 

Results 415 

Overall Changes in Vegetation Distributions and Composition 415 

Logging Impacts 428 

Fire and Vegetation Change 436 

Vegetation Changes in the Areas Affected by the Suwannee River 

Sill 438 

Discussion 465 

5 FIRE IN OKEFENOKEE SWAMP 483 

Introduction 483 

Methods 488 

Wildfires and Prescribed Fires 488 

Wildfire Occurrences and Vegetation Types 491 

Results 492 

Fire Sizes, Frequencies, and Causes 492 

Vegetation Changes Where Fires Occurred 505 

Vegetation Changes Regardless of Fire Occurrence 514 

Wildfire and Logging 530 

Recurrence of Fires 53 1 

Fire Occurrence and Water Levels 532 

Fire and the Suwannee River Sill 541 

Discussion 555 

6 RELATIONSHIPS OF OKEFENOKEE SWAMP VEGETATION 
DISTRIBUTIONS AND THE HYDROLOGIC ENVIRONMENT 564 

Introduction 564 

Methods 568 



vm 



Vegetation Sampling 568 

Preparation of Hydrologic Data 574 

Analysis of Vegetation Data 579 

Results 585 

Species' Environments 585 

Species' Environments and Modeled Hydrologic Changes 681 

Discussion 681 

Species' Associations and the Hydrologic Environment 681 

Vegetation Changes Due Sill Impoundment Effects 697 

7 RESPONSE OF THE OKEFENOKEE SWAMP SEED BANK TO 

ALTERATIONS IN THE HYDROLOGIC ENVIRONMENT 702 

Introduction 702 

Methods 707 

Seed Bank Sampling 707 

Analysis of Seed Bank Emergence Data 712 

Results 713 

Species' Responses 713 

Trends in Response to Hydrologic Conditions 732 

Discussion 744 

Wetland Seed Bank Composition and Vegetation Community 

Dynamics 744 

Effects of the Suwannee River Sill on the Okefenokee Swamp Seed 

Bank 757 

8 SUMMARY AND CONCLUSIONS 760 

Human Activity in the Okefenokee Swamp 761 

Okefenokee Swamp Hydrology 764 

Okefenokee Swamp Vegetation 766 

The Okefenokee Swamp Landscape 768 

APPENDIX A SUWANNEE RIVER SILL AUTHORIZATION BY CONGRESS . 770 

APPENDLX B COMPUTER MODEL CODE FOR HYDRO-MODEL 772 

APPENDIX C VEGETATION TRANSECT LOCATIONS 815 

LIST OF REFERENCES 819 

BIOGRAPHICAL SKETCH . : 835 



IX 






LIST OF TABLES 

lahie page 

1-1. Human-caused manipulations of Okefenokee Swamp vegetation and 

topography occurring during the past 150 years 29 

2-1. Water level recorder elevations and staff corrections, operating period, and 

precipitation gauge locations 40 

2-2. Summary parameters of water level recorders installed at Okefenokee National 
Wildlife Refuge during 12-5-1979 through 6-15-1995. Elevations are in 
meters above mean sea level. Basin delineation is discussed in the Swamp 
Basin Delineation and Characterization section 66 

2-3. Summary of water level and precipitation recorder performance during 

12-5-1979 through 6-15-1995 at Okefenokee National Wildlife Refuge 68 

2-A. Summary of water level recorder performance prorated to the initial operating 

period (1054 days) 72 

2-5. Best correlation pairs and regression equations used to estimate missing water 

level recorder data during 1941-1995 75 

2-6. Daily average precipitation estimated with measurements made daily, and 
approximated with biweekly or monthly calculations of daily averages 
during 31 March 1992-3 July 1995 88 

2-7. Stratum and station variances and covariances of daily precipitation estimates, 
averaged by day, biweekly, and monthly during 31 March 1992-3 July 1995. 
Symbology is defined in the chapter text 90 

2-8. Relative, spatial, and total variances within the precipitation gauge network, 
network accuracy with various recorder densities, and suggested strata 
allocation of precipitation gauges 95 



2-9. Best correlation pairs and regression equations used to estimate missing 

precipitation recorder data for use in HYDRO-MODEL, during 1930-1993 .109 

2-10. Regression relationships used to estimate river and creek outflow rates from 

Okefenokee Swamp during 1930-1993 1 14 

2-11. Estimating creek flow into Okefenokee National Wildlife Refuge using water 

depth estimates from creek staffs 116 

2-12. Regression relationships used to estimates water depths at staffs in 

northwestern creeks from flow measurements at the Suwannee River-Fargo 
gauge 117 

2-13. Monthly latitude adjustment to account for seasonal radiation in calculation 

of Thornthwaite's PE (from Thornthwaite (1948)) 118 

2-14. Regression equations used to estimate missing daily maximum air 

temperature at NOAA weather stations around Okefenokee National 

Wildlife Refuge 119 

2-15. Comparison of flow rates measured at the Suwannee River (Fargo) and the St. 
Marys River (Moniac) gauges before and after construction of the Suwannee 
River Sill, during 1930-1993 132 

2-16. Comparison of pre-sill and with-sill biweekly total precipitation volumes at 

SCFSP and SCRA during 1930-1995 134 

2-17. Comparison of evapotranspiration (ET) estimates in the Okefenokee Swamp 

area before and after Suwannee River sill construction, 1930-1993 137 

2-18. Comparison of SCFSP and SCRA water surface elevations above mean sea 
level (AMSL) before and after construction of the Suwannee River sill, 
1941-1995 141 

2-19. Differences in mean monthly precipitation among decades at SCFSP and 
SCRA, and 95% confidence intervals. No differences were significant at 
a < 0.05. Data were log-normalized before comparisons were made 143 

2-20. Differences in mean biweekly water surface elevation among decades at 
SCFSP and SCRA, and 95% confidence intervals. Differences marked 
with * are significant at a < 0.05 147 



XI 



2-21. Elevations and locations of benchmarks established in the Okefenokee 

Swamp National Wildlife Refuge and perimeter, and peat and sand surface 
elevations above men sea level (AMSL) at each site 152 

2-22. Peat thickness values used to estimate peat depth by vegetation type, to 

supplement the coverage of estimated peat depths where data gaps exist 165 

2-23. Composition and area of classes in the Okefenokee Swamp satellite image 

classification 179 

2-24. Vegetation species found in ground-truthed sites used in the satellite image 

classification 182 

2-25. Error matrix for the 1 1 -class satellite image classification, within 10 pixels 
(100 m) of ground truth sample point location. Rows are reference data; 
columns are classification data. Cell values are number of sample points ... 184 

2-26. Error matrix for the 13-class satellite image classification, within 10 pixels 
(100 m) of ground truth sample point location. Rows are reference data; 
columns are classification data. Cell values are number of sample points. .186 

2-27. Error matrix for the 17-class satellite image classification, within 10 pixels 
(100 m) of ground truth sample point location. Vegetation class number 
refers to Table 2-23. Rows are reference data; columns are classification 
data. Cell values are number of sample points 188 

2-28. Error matrix for the 22-class satellite image classification, within 10 pixels 
(100 m) of ground truth sample point location. Vegetation class number 
refers to Table 2-23. Rows are reference data; columns are classification 
data. Cell values are number of sample points 190 

2-29. Categorical Kappa coefficients (KJ, user's and producer's accuracies for 
classes with > 18 ground-truthed sites in the 1 1-, 13-, 17-, and 22-class 
classifications within 10 pixels (100 m) of sample location 195 

2-30. Class confusions in the 1 1-, 13-, 17-, and 22-class classifications within 10 
pixels (100m) of sample location, for classes with user's accuracy <80%. 
The class with the most frequent error is underlined 197 

2-31. Error matrix of classes in the swamp-islands-and-uplands map that are 
combined with those of the 17-class map. Cell values are number of 
samples re-classified in the swamp-islands-and-uplands classification from 
classes in the 17-class map 200 



xn 



3-1. Manning's roughness coefficients used in HYDRO-MODEL to adjust surface 

water flow rates over various substrates (adapted from Ward (1996)) 219 

3-2. Best HYDRO-MODEL settings and check data format for stations in 

Okefenokee Swamp during 1941-1993 model simulations 260 

3-3. Comparison of check station data and best model output, 1980-1993 268 

3-4. Summary statistics of recorder data and model output at check stations during 

1980-1993 271 

3-5. Water depth ranges for hydroperiod group delineations 282 

3-6. Comparisons of changes in water depths and hydroperiod group frequencies in 
with-sill and no-sill model simulations, 1941-1993. Stations with poor 
with-sill model versus recorder agreement in 1980-1993 are omitted 283 

3-7. Comparison of changes in growing season and non-growing season 

hydroperiod group frequencies in with-sill and no-sill model simulations, 
1941-1993. Stations with poor with-sill versus recorder agreement in 
1980-1993 are omitted 315 

3-8. Comparison of water depths and changes at the south Sill Gate and other check 
stations throughout the Okefenokee Swamp during 1980-1993 with-sill, 
no-sill, and no-outflow model simulations 318 

3-9. Vegetation changes occurring in areas affected by the sill and burned during 

1855-1993 365 

3-10. Composition of vegetation during 1952, 1977, and 1990 throughout the 

Okefenokee Swamp, the floodplain sill impoundment impact area, and the 
Cypress Creek watershed area. All calculations were made with 6-class 
vegetation maps with a minimum mapping unit of 320 m; comparison areas 
for the sill area of impact (AOI) are clipped to match the area interpreted 
from 1952 photography, and reported values are % of the vegetation in each 
category inside and outside the sill AOI during the specified interval 371 

3-11. Rate of change in vegetation composition during 1952-1977 and 1977-1990 

throughout the Okefenokee Swamp, the floodplain sill impoundment affected 
area, and the Cypress Creek watershed area. All calculations are made 
with 6-class vegetation maps with a minimum mapping unit of 320 m; 
comparison areas for the sill area of impact (AOI) are clipped to match the 
area interpreted from 1952 photography, and reported values are % of the 



xni 



vegetation category change occurring in the specified interval. Overall 
change refers to that occurring in the entire swamp, including the sill AOI, 
during that interval. Bare Ground-Urban and Open Water classes were not 
interpreted in the 1977 map, and are omitted from these comparisons 373 

4-1. Sources of pre-logging survey notes used to create the pre-logging vegetation 

map of Okefenokee Swamp 396 

4-2. Vegetation class descriptions and merges created for comparing maps of 

Okefenokee Swamp vegetation distributions during 1990, 1977, 1952, and 
before logging occurred (1850-1890) 399 

4-3. Date groupings for wildfire map sets and for vegetation distribution 

comparisons 414 

4-4. Okefenokee Swamp vegetation composition estimated from pre-logging 

surveys conducted during 1850-1890 417 

4-5. Okefenokee Swamp vegetation composition estimated from an 1 1 May 1990 

SPOT satellite image 419 

4-6. Okefenokee Swamp vegetation composition estimated from 1952 black and 

white aerial photography 421 

4-7. Estimated Okefenokee Swamp vegetation composition compiled from 1977 
color-infrared photography interpreted by McCaffrey and Hamilton (1980) 
with a minimum mapping unit of 320 m 422 

4-8. Landscape level vegetation changes occurring in Okefenokee Swamp during 
1 850-195 1 . Minimum mapping unit for the comparison is 240 m. Reported 
values are % of the vegetation class in 1850 occurring in the specified class 
in 1952 425 

4-9. Landscape-level vegetation changes occurring in Okefenokee Swamp during 
1952-1977. Minimum mapping unit for the comparison is 320 m. Reported 
values are % of the vegetation class in 1952 occurring in the specified class 
in 1977 426 

4-10. Landscape level vegetation changes occurring in Okefenokee Swamp during 
1977-1990. Minimum mapping unit for the comparison is 320 m. Reported 
values are % of the vegetation class in 1977 occurring in the specified class 
in 1990 .427 



xiv 



4-11. Estimated composition of logging tramline areas before logging occurred, 

recorded in surveys conducted during 1850-1890 429 

4-12. Estimated composition during 1952 of areas previously logged 430 

4-13. Estimated composition during 1977 of areas previously logged 431 

4-14. Estimated composition during 1990 of areas previously logged 432 

4-15. Proportions of the entire swamp and logged areas that remained in persistent 
vegetation types between intervals, and the predominant type of 
replacement where changes occurred during 1952-1977 and 1977-1990 434 

4-16. Vegetation changes occurring during 1952-1977 in areas logged during 

1890-1942. Minimum mapping unit for the comparison is 320 m. Values 

are % of the vegetation class in 1952 occurring in the specified class 

in 1977 435 

4-17. Vegetation changes occurring during 1977-1990 in areas logged during 

1890-1942. Minimum mapping unit for the comparison is 320 m. Values 

are % of the vegetation class in 1977 occurring in the specified class 

in 1990 437 

4-18. Vegetation types that burned after 1 855 and before 1952, and the types of 

vegetation that occurred in the burned areas in 1952 439 

4-19. Logging tramline fuel load estimates 440 

4-20. Fuel load composition for fires occurring during 1855-1951, 1952-1976, 

and 1977-1990 441 

4-21. Proportion of wildfires in logged and unlogged tramline areas 442 

4-22. Vegetation types that burned during 1954-1955, and the types of vegetation 

that occurred in the burned areas by 1977 443 

4-23. Vegetation types that burned after 1955 and before 1990, and the types of 

vegetation that occurred in the burned areas in 1990 444 

4-24. Vegetation changes occurring during 1952-1990 in the river floodplain area 
most likely affected by the sill's impoundment and in the Cypress Creek 



xv 



watershed area. Minimum mapping unit for the comparison is 240 m. 

Values are % of the vegetation class in 1952 occurring in the specified 

class in 1990 446 

4-25. Vegetation changes occurring during 1952-1990 in the floodplain area 

most likely affected by the sill's impoundment and in the Cypress Creek 
watershed area. Classes from the 1 990 map have not been grouped; values 
are % of the vegetation class in 1952 occurring in the specified class of the 
ungrouped map in 1990. Minimum mapping unit for the comparison 
is 240 m 448 

4-26. Vegetation changes occurring during 1977-1990 in the floodplain area most 
likely affected by the sill's impoundment effects and the Cypress Creek 
watershed area. Neither map consisted of grouped classes; values are % of 
the vegetation class in the ungrouped 1977 map in the specified class of the 
ungrouped map in 1990. Minimum mapping unit for the comparison 
is 240 m 452 

4-27. Vegetation changes occurring during 1952-1977 in the floodplain area 
most likely affected by the sill's impoundment effects and the Cypress 
Creek watershed. The 1977 map did not consist of grouped classes; values 
are % of the vegetation class in the 1952 map in the specified class of the 
ungrouped map in 1977. Minimum mapping unit for the comparison 
is 320 m 460 

4-28. Vegetation changes occurring in Okefenokee Swamp during 1855-1990. 
Values are % of the prelogging vegetation in the specified class in each 
class during 1990. Minimum mapping unit for the 1990 map is 10 m; 
interpretation of the prelogging survey notes is on a much greater scale, 
from summarization of narratives and observations 466 

4-29. Vegetation changes occurring in Okefenokee Swamp during 1977-1990. 
Values are % of the vegetation from the specified 1977 class changing to 
the specified vegetation type by 1990. Minimum mapping unit for the 
maps is 320 m 472 

5-1. Fuel models used in fuel load calculations, from Anderson (1982) 493 

5-2. Summary of wildfires in the Okefenokee Swamp National Wildlife Refuge 

area, 1855-1993 495 

5-3. Vegetation in 1990 in areas that burned during 1954-1955 511 



xvi 



5-4. Vegetation that burned during 1990-1993 512 

5-5. Types of vegetation changes in 1952-1977, and their proportions in area 

burned and not burned during 1855-1952, 1952-1977, and 1977-1990 515 

5-6. Types of vegetation changes in 1977-1990, and their proportions in area 

burned and not burned during 1952-1955, 1977-1990, and 1990-1993 521 

5-7. Composition of logging harvest during 1909-1927, from Hopkins (1947) and 

Izlar (1984) 530 

5-8. Proportion of burned areas that repeatedly burned 533 

5-9. Spearman rank order correlation comparisons (r s , P) of wildfire size, water 

depths, and wildfire cause, for wildfires occurring during 1941-1993 534 

5-10. Wildfires occurring in the Suwannee River floodplain and Cypress Creek 
watershed areas affected by the sill, water level conditions when the fires 
ignited, and the water level condition that would have been required to 
create impounded surface water and arrest the spread of these fires 551 

6-1. Structural zone types recognized along sampled topographic/hydrologic 

gradients in Okefenokee Swamp 572 

6-2. Recorders and nearest survey benchmarks used to estimate water surface 

elevations at vegetation transects during 1960-1995 576 

6-3. Inundation depth classes defined for analysis of species occurrence in 

hydrologic environments 578 

6-4. Hydrologic environments during 1962-1995 of species occurring in vegetation 
sample plots during 1993-1994. Water depth conditions (DC) are described 
in the table footnote 593 

6-5. t-test (mean water depth) and Wilcoxon rank-sum (percent of interval in each 
depth class) comparisons of hydrologic environments during 1962-1995 
where species were present and absent in vegetation sample plots during 
1993-1994 607 

6-6. Significant parameters in the species-environment multiple regression models 

using all herbaceous and woody species sample data (logit-transformed) 
regardless of species presence 637 



xvn 



6-7. Significant parameters in the species-environment multiple regression models 
using all herbaceous and woody species sample data regardless of species 
presence, when inundation depths are shallow (0 < depth < 0.30 m) 640 

6-8. Significant parameters in the species-environment multiple regression models 

using understory, shrub, and tree sample data where species are present 643 

6-9. Significant parameters in the species-environment multiple regression models 
using herbaceous and woody species data where species are present and 
inundation depths are shallow (0 < depth < 0.30 m) 646 

6-10. Associations of vegetation species based on average and standard deviation of 
daily water depths measured or estimated at transect sample sites during 
1962-1995. Values in column headings are the ranges of average water 
depths + the ranges of standard deviations 648 

6-11. Associations of vegetation species based on similar 3-dimensional plots of 
modeled species occurrence and daily depth-inundation duration 
relationships at transect sample sites during 1962-1995 679 

6-12. Predicted changes in biweekly water depth range (from depth classes) and 

variability by swamp region, predicted by the swamp hydrology model with 

sill removal (summarized from Figure 3-18) 682 

7-1 . Species germinating in the seed bank samples, and their distributions among 

areas, seasons, and treatments 714 

7-2. Standing vegetation species absent from the Okefenokee Swamp seed bank 

samples, but present in plots of established vegetation 719 

7-3. Average number of species in the established vegetation and in the seed bank 

from structural zones throughout Okefenokee Swamp 723 

7-4. Modeled parameters and their significance in predicting responses of seed 

bank species to area, structural zone type, and treatment 726 

7-5. Hydrologic environments where seed bank species are found in Okefenokee 
Swamp, and areas of maximum species abundances in seed bank samples 
and established vegetation 733 

7-6. Effects of season on response of seed bank samples (counts) collected from 

hydrologic zone types and areas of Okefenokee Swamp 736 



xvm 



7-7. Germination and dispersal characteristics of Okefenokee Swamp seed bank 
species, summarized from field observation, seed bank samples, 
Porcher (1995), and Conti and Gunther (1984) 738 









xix 



LIST OF FIGURES 
Figure page 

1-1. The four ecosystem functions and their relationships to the amount of stored 

capital and degree of connectedness, from Holling (1986) 8 

1-2. Interactions of temporal and spatial scales and the processes shaping a 

landscape and its components (adapted from Holling (1992)) 10 

1-3. Interactions of disturbances and system potential energy levels leading to 

instability and restructuring (adapted from Forman and Godron (1986)) 12 

1-4. Location of the Okefenokee Swamp and Okefenokee Swamp National 

Wildlife Refuge 21 

1-5. Hypothetical community changes occurring with peat accumulation in the 
absence of disturbance in Okefenokee Swamp (adapted from 
Hamilton (1982)) 26 

2-1. Water level and precipitation recorder locations in the Okefenokee Swamp 

during 1941-1995 38 

2-2. Daily water surface elevation above mean sea level (AMSL) during 1941 -May 

1995 recorded at locations in the Okefenokee National Wildlife Refuge, GA. . 42 

2-3. Recorder distribution for daily measurement of precipitation at 1 1 stations in 

Okefenokee Swamp 98 

2-4. Recorder distribution for daily measurement of precipitation at 14 stations in 

Okefenokee Swamp 99 

2-5. Recorder distribution for biweekly measurement of precipitation at 1 1 stations 

in Okefenokee Swamp 100 

2-6. Recorder distribution for biweekly measurement of precipitation at 14 stations 

in Okefenokee Swamp 101 

xx 



2-7. Recorder distribution for monthly measurement of precipitation at 1 1 stations 

in Okefenokee Swamp 102 

2-8. Recorder distribution for monthly measurement of precipitation at 14 stations 

in Okefenokee Swamp 103 

2-9. Recorder locations for daily measurement of air temperature and surface water 

inflows in Okefenokee Swamp 113 

2-10. Variance contours for water depths recorded daily at locations in the 

Okefenokee Swamp during 1992-1995 121 

2-11. Water level recorder locations and hydrologic basins in Okefenokee 

Swamp 123 

2-12. Trends in water level fluctuations in the Okefenokee Swamp hydrologic 

basins 124 

2-13. Monthly average evapotranspiration estimated in the Okefenokee Swamp 

area during 1930-1993 126 

2-14. Average monthly precipitation estimated at sites in the Okefenokee Swamp 

area during 1930-1993 127 

2-15. Average daily inflow estimated for northwestern creeks entering Okefenokee 

Swamp during 1930-1993 128 

2-16. Average daily outflow estimated for creeks and rivers exiting Okefenokee 

Swamp during 1930-1993 129 

2-17. Average monthly precipitation and water surface elevation reported by 5-year 

intervals at SCFSP, during 1941-1994 149 

2-18. Locations surveyed and extracted from USGS 1994 1:24,000 topographic 

maps for development of the Okefenokee Swamp topographic surface 161 

2-19. Correction surface used to adjust the swamp topography map, generated from 

the difference between the actual and interpolated surface elevation data 162 

2-20. Peat and sand surface topography in Okefenokee Swamp. Darker areas are 

lower in elevation above mean sea level 163 

2-21. Estimated thickness of peat over the basement sands in Okefenokee 

Swamp 166 

xxi 



2-22. Estimated sand surface elevation above mean sea level under the surface 

peat in Okefenokee Swamp 167 

2-23. Elevation gradients recorded along transects in Okefenokee Swamp prairies . . 168 

2-24. Creeks, rivers, lakes, and canoe trails where surface water flow occurs in the 

Okefenokee Swamp 1 70 

2-25. Surface water drainage patterns and underlying topographic gradients in 

Okefenokee Swamp 171 

2-26. Merging 10 m pixel panchromatic and 20 m pixel multispectral imagery to 

create a multispectral image with 10 m pixel resolution 175 

3-1. Processing area for the Okefenokee Swamp HYDRO-MODEL 211 

3-2. Flowchart of Okefenokee Swamp HYDRO-MODEL components 212 

3-3. Neighborhood search in HYDRO-MODEL to determine direction and amount 

of water to move in each cell and time interval 214 

3-4. HYDRO-MODEL menu interface for setting user-defined parameters 215 

3-5. Locations of zones used in HYDRO-MODEL to distribute inflowing water 

across the landscape 217 

3-6. Location of zone used in HYDRO-MODEL to extract outflowing water from 

the Suwannee River floodplain 218 

3-7. Locations of water level recorders used to assess HYDRO-MODEL 

performance 224 

3-8. Estimated topographic surface representing the pre-sill peat surface elevations. 

Dark areas are low in elevation 226 

3-9. Estimated area of impact of the Suwannee River sill on the Okefenokee 

Swamp hydrologic environment during various water level conditions, and 
regions that may experience head reversals when water levels are high 230 

3-10. Estimated recorder data and model output from the "with-sill" and "no-sill" 

simulations for 1980-1993 231 



xxn 



3-11. Estimated recorder data and model output from stations with poor model 

performance in "with-sill" and "no-sill" simulations for 1980-1993 251 

3-12. Inverse-distance- weighted, contoured estimates of increases in average semi- 
monthly water surface elevations (m) at recording stations, attributed to the 
Suwannee River sill during 1980-1993 276 

3-13. Comparison of semi-monthly water surface elevations at Cypress Creek and 
Sapp Prairie under increasing water level conditions in the sill gate area 
during 1980-1993 277 






3-14. Comparison of average, semi-monthly water surface elevations in the 

Cypress Creek watershed under low, average, high, and very high water 

levels in the sill gate area during 1980-1993 278 

3-15. Locations of topographic highs in the Suwannee River floodplain near the 

Suwannee River sill and Cypress Creek 280 

3-16. Comparison of average, semi-monthly water surface elevations in the 

Sweetwater Creek watershed under low, average, high, and very high water 
levels in the sill gate area during 1980-1993 281 

3-17. Changes in most frequent hydroperiod groups during 1980-1993, with sill 

removal. Numbers represent the most frequent with-sill hydroperiod groups 
(see Tables 3-5 and 3-6) versus most frequent no-sill hydroperiod groups. 
Areas with significant change are marked with *. Average semi-monthly 
water depth decrease with sill removal is noted in ( ) 294 

3-18. Changes in hydroperiod depth group frequencies with and without the sill 

during 1941-1993, by decade intervals 295 

3-19. Comparison of semi-monthly water surface elevations at recorder stations 
and the sill gate during 1980-1993, in "with-sill" and "no-sill" model 
simulations. Data are arranged to illustrate the change in water levels 
at the selected stations, relative to increasing water depth at the sill 335 

3-20. Estimated average biweekly flow rates at Suwannee River (Fargo) and St. 

Marys River (Moniac) during 1980-1993 345 

3-21. Locations of topographic change in the Suwannee River floodplain within the 

Okefenokee Swamp and southwest of the Suwannee River sill 347 

3-22. Inverse-distance-weighted, contoured estimates of increases in average semi- 
monthly water surface elevations (m) at recording stations, attributed to 

xxiii 



Suwannee River outflow retained in the swamp during 1980-1993 model 
simulations 349 

3-23. Manipulations of estimated evapotranspiration rates and responses of the 

model at recorder stations during 1980-1993 351 

3-24. Areas affected by the Suwannee River sill at various water level conditions, 

and burned by wildfires during 1960-1993 362 

4-1. General effects of fire and logging disturbances on Okefenokee Swamp 

vegetation types, adapted from Hamilton (1982) 387 

4-2. Autogenic succession and conditions that drive succession in the Okefenokee 

Swamp, adapted from Hamilton (1982) 389 

4-3. Locations of logging railroads (tramlines) and buffer used to approximate 

logged area 394 

4-4. Estimated pre-logging (1850-early 1900s) vegetation in Okefenokee Swamp, 

and approximate routes of 19th and 20th century surveyors 397 

4-5. Photo interpretation results of Okefenokee National Wildlife Refuge 

vegetation during 1952 409 

4-6. Scanned, transformed version of map created by McCaffrey and Hamilton 

(1980) of Okefenokee Swamp vegetation during 1977 412 

4-7. Vegetation distributions delineated in a classification of 1990 SPOT satellite 

imagery 413 

4-8. Areas of vegetation change occurring during 1952-1990, and regions where 

the swamp hydrologic environment has been affected by the sill 447 

4-9. Dominant successional processes occurring in the Okefenokee Swamp 

landscape during 1977-1990. Although all vegetation types were present at 

both times, dominant community types at each time changed between years, 

as indicated by types in rectangles (1977) and ellipses (1990). Arrows and 

numbers indicate changes under various conditions. Greatest increase was 

in cypress-bay-blackgum. Mixed blackgum swamp was a minor type in 

both years (modified from Hamilton (1982)) 476 

5-1 . Number and cause of wildfires reported in the Okefenokee National Wildlife 

Refuge area during 1 855-1936 497 



XX1V 



5-2. Total area burned in the Okefenokee National Wildlife Refuge area by 

wildfires during 1855-1936 498 

5-3. Total area burned by wildfires during 1855-1936, excluding the large fires of 

1931-1932 499 

5-4. Total area burned in the Okefenokee National Wildlife Refuge by wildfires 

during 1936-1993 500 

5-5. Number and causes of wildfires reported in the Okefenokee National Wildlife 

Refuge during 1937-1993 501 

5-6. Total area burned by wildfires during 1937-1993, excluding those in 

1954-1955 502 

5-7. Number and causes of wildfires preceding (1855-1936) and following 

(1937-1993) Okefenokee National Wildlife Refuge establishment 503 

5-8. Total area burned and causes of wildfires occurring in Okefenokee Swamp by 

intervals during 1855-1993 504 

5-9. Number and causes of wildfires in Okefenokee Swamp during 1855-1993 506 

5-10. Total area burned by wildfires in Okefenokee Swamp during 1855-1993 507 

5-11. Prescribed burning compartments in the Okefenokee National Wildlife 

Refuge 508 

5-12. Water levels at Stephen C. Foster State Park (SCFSP) and Suwannee Canal 
Recreation Area (SCRA), and total prescribed burning area, during 
1973-1993 509 

5-13. Water levels at Stephen C. Foster State Park (SCFSP) and Suwannee Canal 

Recreation Area (SCRA), and the number of wildfires reported monthly 536 

5-14. Water levels at Stephen C. Foster State Park (SCFSP) and Suwannee Canal 

Recreation Area (SCRA), and area burned by wildfires 542 

5-15. Total area of Okefenokee National Wildlife Refuge burned by prescribed 

fires during 1973-1993 547 

5-16. Total area of Okefenokee National Wildlife Refuge burned by prescribed 

fires and wildfires by intervals during 1855-1993 548 



XXV 



5-17. Locations of wildfires in Okefenokee National Wildlife Refuge, during 
1855-1959 and 1960-1993. Extent of each fire is indicated by a black 
outline 549 

5-18. Hypothesized return frequency, duration, extent, and intensity of drought and 
wildfires in Okefenokee Swamp, and the extent and permanency of 
subsequent vegetation changes 560 

6-1. Locations of vegetation transects sampled during 1993-1994 in Okefenokee 

Swamp 569 

6-2. Schematic diagram of the placement of understory, overstory, and shrub plots, 
and tree belts along a vegetation transect sampled during 1993-1994 in the 
Okefenokee Swamp 570 

6-3. Example interpretation of axes (Graph 1) and curvatures (Graph 2) on 
3-dimensional plots of model-predicted abundances of species with 
flooding depth and duration 583 

6-4. Average daily water depths (1962-1995) for species recorded at Okefenokee 

Swamp sample sites during 1993-1994 586 

6-5. Average daily water depths along vegetation transects sampled in Okefenokee 

Swamp during 1993-1994 591 

6-6. Trends in abundances of all species occurring along an exposure gradient 

during 1993-1994 in Okefenokee Swamp 592 

6-7. Hydrologic conditions where species occurred at greatest abundance 

(90-100% maximum density or percent cover) 621 

6-8. Distribution of sample points in observed and model-predicted relationships 
between species abundance (1993-1994) and inundation depth and duration 
(1962-1995) 650 

7-1. General scheme of seed bank sample collection sites, relative to other 

vegetation sample plots, along transects in Okefenokee Swamp 708 

7-2. Seed bank emergence experiment sample layout in greenhouse with 

continuous swamp water irrigation system 710 



xxvi 



7-3. Counts of species and germinated seeds in seed bank samples, compared to 
species counts in the standing vegetation within the structural zone at the 
collection site. Structural zones are described in Table 6-1 721 

7-4. Vegetation structural zones arranged with increasing duration of deep water 
flooding (> 0.30 m; dotted line). Duration of exposed conditions is plotted 
(solid line), and variance in average daily water depth for each zone type is 
indicated. Species generally associated with each zone type in the seed 
bank and established vegetation are listed in Tables 7-4 and 7-6 756 



xxvn 



Abstract of Dissertation Presented to the Graduate School 

of the University of Florida in Partial Fulfillment of the 

Requirements for the Degree of Doctor of Philosophy 

ASSESSING PATTERNS AND PROCESSES OF LANDSCAPE CHANGE 
IN OKEFENOKEE SWAMP, GEORGIA 

By 

Cynthia Smith Loftin 

December 1998 

Chairperson: Dr. Wiley M. Kitchens 

Major Department: Wildlife Ecology and Conservation 

The Okefenokee Swamp is one of the largest freshwater wetlands in the world. 
Currently protected and managed as a national wilderness area and national wildlife 
refuge, the swamp has a history of human-caused manipulation and modification. The 
swamp landscape is dynamic; vegetation compositions and distributions continually 
change as the hydrologic environments change. These dynamics are driven by natural 
processes such as peat accumulation and wildfire, as well as the artificial manipulations 
of the recent past. 

The Suwannee River sill was constructed following extensive wildfires during 
1954-1955, with the intent of protecting the swamp and surrounding uplands from effects 
of wildfires. During subsequent years, concern was raised that the dam might be 
adversely affecting the swamp ecology by extending periods of inundation, increasing 

xxviii 



water depths, and subsequently affecting swamp vegetation. Delineating the effects of 
the Suwannee River sill on the swamp hydrologic environment and vegetation 
distributions, in the process of exploring relationships among driving functions and 
landscape responses, was a purpose of this dissertation research. 

Data collected at various spatial and temporal scales were examined to identify 
the sill's effects. A water level recorder network was spatially linked with a global 
positioning system survey, and the resultant topographic surface and hydrologic data 
were included in a grid-cell based hydrology model to track water movement throughout 
the swamp. Model simulations illustrated swamp water level fluctuations before and 
after the sill was in place, and predicted recent hydrologic history in the sill's absence, as 
well as sensitivities of swamp hydrology to altered evapotranspiration rates. Model 
simulations also predicted that the sill was affecting about 18% of the swamp area with 
increased inundation depths and durations, and vegetation change attributed to the sill 
was limited to this area. 

Vegetation dynamics were also assessed at several scales, with remote sensing 
techniques, species-hydroperiod descriptions, and seed bank analysis and hydrologic 
manipulation. Current vegetation distributions are artifacts of historic logging and recent 
lack of fire, and also show sensitivity to local hydrologic environments. Inundation 
depth and hydroperiod create hydropatterns that influence species distributions. The 
swamp landscape is an expression of local dynamics, coupled with landscape-level 
processes such as fire, drought, and extensive historic logging occurring at multiple 
temporal scales. 

xxix 



CHAPTER 1 
THE OKEFENOKEE SWAMP AND THE SUWANNEE RIVER SILL 



The Okefenokee Swamp is a 200,000 ha freshwater wetland in Southeast Georgia 
and Northeast Florida. The landscape was relatively undisturbed by American 
explorers and settlers until the end of the nineteenth century, when it was subjected to 
draining, timber harvest, and mining. Protection and preservation of the landscape and 
remaining resources were goals in 1937, when the swamp became part of the National 
Wildlife Refuge system. The Suwannee River Sill, constructed in 1960 across the main 
outflow channel of the Suwannee River where it exited the swamp, was also intended to 
protect and preserve the swamp. Built in response to fires that burned across the swamp 
and into the surrounding landscape during 1954-1955, the sill was to impound water in 
the Okefenokee Swamp to keep similar fires from igniting and burning in the swamp. 
During the 30 years following construction of the sill, refuge managers, biologists, and 
the public exploring the swamp noticed changes in the composition and distributions of 
vegetation communities throughout the swamp. Were the changes indicating that the sill 
was affecting swamp vegetation, or was it undergoing natural successional processes? 
Concern for the health of the swamp ecosystem began to emerge. During 1989 the 
conditions of the sill gate structures were reviewed and found to be unstable, and in need 
of repair. Should the sill gates and impoundment berm be repaired, modified, or 



1 



2 
destroyed? Was the Suwannee River Sill responsible for altering swamp vegetation, or 
were the perceived changes artifacts of the observers' temporal and spatial scales? 
Rather than "protecting" the swamp, was the sill damaging the wetland by disrupting the 
natural hydrologic environment and subsequently the vegetation community dynamics? 

Addressing these questions presents an opportunity to examine the Okefenokee 
Swamp landscape composition and structure, and identify processes that create and 
maintain this structure. Hydrology is a primary driving function of all wetlands, and the 
hydrologic regime, principally hydroperiod, determines wetland type (Mitsch and 
Gosselink 1986). Many wetlands are also shaped by fire, and fire suppression may 
compromise wetland integrity (Mitsch and Gosselink 1986). In many wetland systems 
fire and the hydrologic regime are intricately linked; periodic droughts create conditions 
favorable for burning. Fires occur, potentially altering site environments (e.g., soil 
composition, site elevation, and hydrologic features), and subsequent species 
composition. Alterations of frequencies, intensities, and extent of these processes (fire 
and the hydrologic regime) can modify landscape composition and structure (DeAngelis 
and White 1994). Human activity has disrupted Okefenokee Swamp hydrology and fire 
regimes. Hierarchy theory suggests that the extent of these disruptions depends on the 
organizational level of the swamp ecosystem that is normally affected by these processes, 
and the relative importance of the affected driving function in maintenance of the system 
hierarchy. 

The purposes of this work are to describe the spatial, hydrologic environment of 
the Okefenokee Swamp, to identify changes in vegetation community distributions since 



3 
the sill's construction and their probable causes, and to examine the swamp landscape 
structure and driving functions in the context of hierarchy and succession theories. To 
achieve these goals it was necessary to analyze the swamp vegetation and shaping 
functions from several spatial and temporal scales. Hydrologic monitoring and 
topographic surveying at locations throughout and surrounding the swamp provided data 
for describing the swamp hydrology. Remote sensing and ground truthing provided 
landscape-level vegetation distribution information. Transects across topographic 
gradients provided relational data among species occurrences, hydrologic features, and 
site conditions. Seed bank composition, source, and response to hydrologic regimes 
imposed in controlled greenhouse conditions suggested species germination sensitivities 
and potential responses to changing hydrologic conditions. Wildfire and prescribed 
burning records provided a spatial history of fire to compare with hydrologic and 
vegetation distribution information. Pre-logging surveys implied vegetation distributions 
resulting from natural successions, and logging records, historic aerial photography, and 
recent satellite imagery provided an indication of the extent and duration of logging 
impacts. A spatial hydrology model was used to estimate the spatial extent of the sill's 
influence on the swamp hydrologic environment. And, a geographical information 
system (GIS) was used to identify the spatial relationships of all of these components to 
the sill and to current vegetation community distributions and hydrologic features, 
elucidating the sill's effects on the swamp ecosystem. 

This dissertation is arranged in 8 chapters. The first chapter discusses the 
application of theories of hierarchy, scale, and succession to dynamics in the wetland 






4 
landscape. A description of the Okefenokee Swamp ecosystem, history of human 

influence, and discussion of the history and intended purpose of the Suwannee River Sill 
are also contained in the chapter. Chapter 2 describes the acquisition and management 
of data included in development of the spatial hydrology model (precipitation, 
evapotranspiration, flow, and topography), vegetation distribution and species- 
hydroperiod associations (satellite image classification and accuracy assessment, aerial 
photography interpretation, and hydroperiod calculations), and precipitation and water 
level recorder network accuracy assessments. Swamp hydrologic conditions during the 
period, 1941-1995, are summarized in Chapter 2. Development, implementation, and 
assessment of a swamp hydrology model (HYDRO-MODEL) are detailed in Chapter 3. 
An assessment of the impact of the Suwannee River Sill on the swamp hydrologic 
environment based on the HYDRO-MODEL output is made in Chapter 3. Changes in 
vegetation community distributions since before logging occurred were detected with 
comparisons of pre-logging survey notes, post-logging aerial photography, and satellite 
imagery interpretations; these results are summarized in Chapter 4. Chapter 5 details the 
wildfire and prescribed burning history; interactions of fire history and swamp hydrology 
with vegetation changes are identified. Hydrologic environments associated with swamp 
vegetation species during 1962-1995 are described in Chapter 6, and seed bank 
composition and response to experimentally altered hydrology are detailed in Chapter 7. 
A synthesis of swamp vegetation succession in response to hydrologic alterations, 
logging, and wildfire management, and the role of these functions in shaping the swamp 
landscape is presented in Chapter 8. 



5 
This introductory chapter provides the theoretical basis of this wetland landscape 
analysis. First, the hierarchical structure of the components and processes of a landscape 
and the effect of observer scale on recognizing this organization are addressed. 
Thereafter, the processes that structure the wetland landscape (driving functions of fire 
and hydrology, and the "disturbances" they create), and the system's responses 
(perceived homogeneity and heterogeneity of the landscape, succession, and the 
resiliencies inherent in all systems) are discussed, as are techniques for studying 
landscape change (GIS, spatial modeling). Finally, a description of the Okefenokee 
Swamp environment, and a history of human impacts on the system are presented. The 
chapter concludes with a discussion of the Suwannee River Sill history and the primary 
questions directing this research. 

Ecosystem Flierarchies and Scales 

Landscapes are the expression of multitudes of components and processes 
interacting at varying temporal and spatial scales. A hierarchical arrangement of these 
components and processes is the framework of ecosystem organization (O'Neill et al. 
1989a, 1989b). Ecosystem predictability and stability are dependent on preserving the 
processes and components occurring at multiple spatial and temporal scales that have 
resulted in the expressed system structure (O'Neill et al. 1989, O'Neill 1989, Holling 
1987, 1986, Allen 1987, Urban et al. 1987, Allen and Starr 1982). Hierarchy theory 
(Allen and Starr 1982) recognizes nesting of system functions and properties with finer 



6 
scale. This nesting arrangement provides the framework in which an ecosystem is 
structured (Allen and Wyleto 1983). Interactions may occur among and within levels in 
the hierarchy, and the outcomes of these interactions may be predictable, until a 
disruption in the system's usual processes may lead to development of a system of 
different components. Eventually a new hierarchical framework develops, which may 
not contain the same components and may result in a different but stable system, 
components, and processes; a new stability domain develops, as the system reorganizes 
in response to these changes (Holling 1987, 1986, 1973). For example, each individual 
plant in a wetland cycles through a period of germination, growth, reproduction, and 
senescence, and many individuals are in various parts of this cycle at any time. An 
individual may live its entire life under a fairly constant, predictable environment. Some 
individuals and species in some years, however, will encounter limiting environments, 
such as extreme drought, which may eliminate them from the standing vegetation in the 
landscape. In the altered environment other species find the conditions suitable for their 
growth and survival. Thus, environmental modification (e.g.,long-term drought) can 
result in changes in species composition and changes driven by processes occurring at a 
scale and hierarchical level greater than the individual, i.e.,the wetland or the region. A 
new hierarchy results, with different species and environmental conditions, and possibly 
driven by different controlling functions. The original species may be present in the seed 
bank, however, and may again become part of the standing vegetation, given a return to 
conditions suitable for germination and maturation. The sustainable system depends on 
this type of adaptive cycle, whereby the system develops, becomes stable, undergoes 



disruption, reorganizes, and has the potential to return to its original design or reorganize 
into another level in the hierarchy (Figure 1-1) (Holling 1987, 1986). A heterogeneous 
landscape indicates that a hierarchy of processes is operating at different spatial and 
temporal scales (O'Neill 1989). A connectivity among levels of the hierarchy that 
responds to variabilities of the systems' processes and components at various spatial 
extents is essential to the system's sustainability (Holling 1995, Allen and Wyleto 1983). 

Stability perceived at the landscape level is a function of dynamics acting locally 
at a small scale, as well as at a greater extent. Thresholds exist whereby changes in the 
system components and processes disrupt the function of the system; these changes 
might be at any level of the system's hierarchy. The effect on the landscape could be 
innocuous, such as elimination of single individuals from a large population of great 
extent, or could result in the restructuring of the entire system by removing entire species 
or communities. A system's complexity is defined by the boundaries of the multiple 
levels and the interacting relationships among them. In general, large structures and low 
frequency processes occupy high levels of hierarchy and affect multiple layers of the 
system, whereas small structures and high frequency processes are low in the hierarchy 
and have limited effects (Ahl and Allen 1996). Predictability in the behavior of the 
system and recognition of the underlying, hierarchical model of the system's design and 
controlling processes result from observation at multiple levels of organization or scales 
(Ahl and Allen 1996). 

The study of ecological processes requires selection of appropriate data resolution 
and extent. Perception and interpretation of the landscape, its patterns, driving 



o 

3 

1 



t 



- 
< 

< 

u 

Q 

- 

OS 

o 

H 
t/3 



t 



w 

13 



4 Renewal 

* Accessible carbon, 
nutrients, and energy 



1 Exploitation 



* r-strategy 

* Pioneer 

* Opportunist 



Conservation 2 



* K-strategy 

* Climax 

* Consolidation 



* V * 



"Creative" 


3 


Destruction 




♦Fire 




♦Storm 




♦Pest 




♦ Senescence 






ORGANIZATION 
CONNECTEDNESS 




Figure 1-1. The four ecosystem functions and their relationships to the amount of stored 
capital and degree of connectedness, from Holling (1986). 






processes, composition, and changes are dependent on the scale of observation (Holling 
1992, O'Neill et al. 1989a, 1989b, Forman and Godron 1986, Allen and Wyleto 1983) 
(Figure 1-2). For example, spatial scale is important to determining perceived effects of 
disturbance in a landscape; what appears as disruption at a local scale may actually be 
maintenance of the landscape mosaic at a regional scale (Risser 1991). Fire may 
maintain the landscape mosaic by changing the distributions of communities in 
landscape, but it may be a disturbance if it completely eliminates the potential return of 
the species in the landscape. Fine-scale measurement narrows scope and restricts extent; 
as data resolution becomes more coarse and extent increases, scope increases and the 
range of potential values for a particular landscape variable and its controlling processes 
also increase. The observer defines a measurement scale when the objectives are stated; 
a particular question posed by an observer defines the scale and hierarchy of interest. 
Selecting an inappropriate scale may lead to misinterpretation of patterns and driving 
processes. Identifying the appropriate data scale for detecting structure in a landscape 
and determining the processes that shape that landscape are fundamental to recognizing 
the landscape's hierarchical organization. Assessment of the effects of human-induced, 
landscape-level perturbations on a complex ecosystem requires integrating information 
from multiple disciplines and data resolutions. Analyses of system response to 
perturbations must address these interacting components, processes, and scales 
(Meentemeyer and Box 1987). Understanding how landscape pattern recognition 
correlates with scale facilitates compilation of information across scales (Farmer and 



10 



CO 

o 

o 
o 
o 



L. 
CO 
0) 

o 
o 
o 



W 

CO u _ 

© co £ 

o >> CO o CO 

° ° ® S Q 

T" T~ >- 



3 

o 



0) 

3 




w C Q. Q) t- 

§| 8 g 8 

Q Q o -3 O d. 





CO 


a _ s 


00 


N S 


c 
'5. 


4-> 


■ 


o> 


£ 


E 


V) 
V3 


o 


6 


^ 


VI 


o * 


VI 



? 


u 

2 


-J 


a 




o 




<v | 


c 


CO 


a 




M 




J£ 






t 


« 




•s 




I 




Vl 




T3 




a 




1 




a 




E ~ 




* 2 




<*- £ 




O & 








C/3 v - 




1 J 




■*■» • •■* 




U S5 




g o 




i> x 








1 E 




o 




<n <~ 




J. T3 




, 1> 




1) ■« 








- 

MT3 




tz A 



11 

Adams 1991, Musick and Grover 1991). Viable management of an ecosystem ultimately 
depends on recognition of and continued interactions of these multiple-resolution 
components (Soule 1985). 

Driving Functions 

Spatial pattern in a landscape is the expression of interactions of current or 
historic processes and system components, and may determine the structure and function 
of the future landscape (Forman and Godron 1986). These determinant or controlling 
processes are the driving functions of the landscape. The dynamics of fire and hydrology 
in a wetland are driving functions that result in a shifting mosaic of communities across 
the landscape in various stages of succession, with current species composition reflecting 
a combination of inter- and intraspecific interactions and the recent driving 
environmental influence. Current compositions and distributions of communities in the 
landscape offer indications of the driving processes influencing the landscape in the past. 
Metastability (equilibrium) of the landscape may increase in the absence of disturbance 
(e.g., fire), allowing the successional sequence to progress and requiring a greater degree 
of disturbance over time to disrupt the equilibrium (Forman and Godron 1986) (Figure 1- 
3). Distribution of species in the landscape, and knowledge of species' sensitivities to 
environmental conditions may elucidate the processes exerting greatest control on the 
landscape structure. 







12 






00 






_c 






0? 






0) 








c 






B « c 


"O 






|| 


§ 






•2 1 


£ 






JZ o 








O "D 


15 




— ^k. 


■ >, 


2 




1 ^ 


dto 
ibilit 


o 




/ 


25 & 

.= CO CO 


00 

c 












./\. .^^^ 


© ■= s 


_o> 




TT ^^ 


u o • 


1/5 






energy 
and rea 
otential 


> 

c 
5 




11 wl 


o £ i» 


c 
6 




_j S^^^ ^t 


c .5 a; 


^2 




^^^^ E 


CO «- JC 


'B 




ibrium 

uilibriu 
in 


turb 
uilib 
fhig 


u 

a 


*c 


isturbance 
nstable Equili 
letastable Eqi 
tability Doma 


Increasing dis 

an unstable eq 

o 


E 
6 

I 

n 

"S ^ 
5 ft 

C3 v£3 




^^^ JS Q 3 * <» 


sturbances 
odron(198< 








>» 


■6 O 




erg 
lain 






c c 


i B 




© o 


•2 i 
u £ 




S3 a 


- O 




8 ? 


£ E 




■£ c 


5 




o •- 


r^' IB 




a. 


Figure 1- 
(adapted 



13 
Hydrology is a driving force in shaping the function and structure of wetlands. 
Hydrologic conditions such as water depth, flood duration and frequency, and water flow 
patterns influence both physical and biotic processes. Primary productivity, 
decomposition of plant material, nutrient cycling and availability, and vegetation 
composition are to some degree controlled by hydrology. Temporal constancy of a 
wetland' s hydrology may be the dominant factor determining its biotic composition 
(Mitsch and Gosselink 1986). Increased flood duration may lower plant species richness 
as flood-intolerant species are eliminated, while decreased flooding or more frequent 
drawdowns may promote nutrient cycling, decomposition, and primary productivity. 

Modifications of a wetland's hydrologic regime can alter the species 
composition, distribution, and productivity. Prediction of vegetation changes must 
consider relationships between hydrology and ecophysiology of individual species (Leek 
et al. 1989). Some plant species respond to flooding with inhibition of seed production 
and germination, retarded shoot, cambial, and root growth, arrested reproductive growth, 
and death. Wetland plants have mechanisms to acclimate to stresses of inundation, such 
as reduced gaseous exchange in a flooded environment. Formation of adventitious roots, 
aerenchyma tissue, hypertrophy of stem lenticels, secondary roots, and formation of 
knees or pneumatophores may be structural changes occurring in response to flood stress, 
to increase exchange of oxygen and waste products (Kozlowski 1984a, 1984b). A plant's 
age, duration of flooding, and the nature of the floodwater influence its response. If 
flooding persists species are replaced by flood-tolerant wetland species (Kozlowski 
1984a, 1984b), usually resulting in a community with lower species diversity. 



14 
Elimination of flooded conditions may encourage return of the displaced species. 
Deuver (1988) and Duever et al. (1987) demonstrated that hydroperiod is a determinant 
of the distribution and composition of freshwater wetland communities in Florida. 
Disruptions in a wetland's hydrologic environment could lead to landscape-level 
structural changes in the wetland. Hydrology is a high-level controlling process in the 
wetland system hierarchy; its disruption could lead to a new hierarchical framework for 
the wetland system. 

In addition to hydrology, fire is an environmental force that may shape a wetland 
over time. In southern Stillwater swamps fire plays an important function in sculpting 
ecosystem structure (Ewel 1990). Occurrence of these fires is affected by seasonal 
cycling of hydrology and periodic droughts. During drier periods fires combust living 
and standing dead vegetation, litter, and normally saturated layers of peat. Although 
rapid community replacement might occur when mild fires leave many species alive to 
resprout, intense fires can eliminate all of the standing vegetation and possibly the peat. 
The seed bank in the exposed peat or sediment becomes the initial regenerative source, 
supplemented by seeds dispersed from adjacent sources. At this time species that can not 
germinate in inundated conditions can become established, provided the burned peat 
surface remains exposed. A similar response may occur when normally submerged 
substrate is exposed by drought. If new species establishing during dry periods are not 
tolerant of inundated conditions, they may be eliminated when water levels rise. With 
artificially extended inundation and limited fire disturbance, establishment of 
inundation-tolerant species occurs, slowly changing the local vegetation community 



15 
composition and structure and ultimately the landscape to a long-hydroperiod system. 
Intentional removal of fire while maintaining natural hydrologic processes can also 
restructure the landscape, by replacing fire-dependent species with those intolerant of 
fire. 

Disturbances. Heterogeneity, and Succession 

Changes in landscape structure may result from influences of the driving, 
functional processes that historically shaped the landscape. Changes may also reflect 
recent alterations in those processes that shape levels in the system's hierarchy. Whether 
changes in landscape structure and composition are perceived as disruptive to the system 
depends on the observer's scale, objectives, the type and intensity of the disturbance, and 
the system's evolution. A disturbance creates unsuitable conditions for some component 
of the system; its effects may be at various scales, favor some species and eliminate 
others, and may be essential in a system's maintenance. A true disturbance is a type of 
disruption absent from a system's evolutionary history (Rapport et al. 1985). Periodic 
fire and drought are driving processes in a wetland where species have evolved under 
their influence. The effects of fire and drought may not be disruptive to the system 
overall; the persistence of some species in the landscape may be dependent on occasional 
fire or drought to improve conditions for germination, remove competitors, or modify 
conditions that promote succession. Removal of these disturbances, which are driving 
functions in many wetlands (e.g.,fire or hydrologic cycling, such as drought-flooding 



16 
cycles), may be more detrimental to a species that evolved in a system maintained by 
periodic disruptions. For example where fire is removed, competitive advantage of fire- 
intolerant species may permit displacement of those dependant on fire. Artificially long 
hydroperiods and excessive water depths resulting from impoundment eliminate 
germination and reduce survival of species not adapted to those conditions. Unnatural 
disruptions on an ecosystem may affect biological diversity and processes with which the 
system evolved, which could ultimately damage the health of the system (Soule 1985). 
Ultimately, an altered stability domain is reached when a system's resilience to these 
perturbations is exceeded (Holling 1995, 1987, 1986). 

Spatial heterogeneity in communities across a landscape reflects species sorting 
in a spatially diverse biotic and abiotic landscape (Milne 1991). Heterogeneity is scale- 
dependent; what appears heterogeneous at one scale is homogeneous at another 
(Meentemeyer and Box 1987). Landscape homogeneity may express a synergy of 
functions, which at a smaller scale appears to create heterogeneity (Meentemeyer and 
Box 1987). Although homogeneous landscapes are thought to enhance the spread of 
disturbance, heterogeneity may also exacerbate effects of disturbance by increasing 
exposure of the landscape interior through percolation (Risser 1991). Change in 
heterogeneity with spatial scale may reflect the functional organization of and shaping 
processes in the landscape (Musick and Grover 1991). A holistic approach to studying 
ecosystem responses recognizes that the effects of disturbance and change may occur 
over varying temporal and spatial scales, and differentially influence individual system 
components. 



17 
Environmental modification and changes in plant community distributions and 
composition in the landscape co-occur. The environmental change might be in response 
to stochastic events, such as fire or extreme weather, or caused by the landscape's 
occupants, such as peat accumulation, chemical soil modification, or shading. A suite of 
species will be adapted to the general conditions of the geographic region, i.e., weather 
and geologic history will determine the potential species pool for the area. Which 
species are present in the standing vegetation will depend on the propagule source, 
competitive interactions among species, and environmental limitations of the site. While 
an individual occupies a site, it gradually modifies the site's characteristics, so that 
conditions become less suitable for itself and more favorable for other species. The site 
will undergo changes in species composition as the physical characteristics of the site are 
modified, eventually altering community structure and ultimately modifying the 
landscape. This change, or succession, in species composition driven by physiographic 
and biotic agents was first described in detail by Cowles (1911). Clements (1916) 
observed specific associations of species occurring in predictable sequences in 
colonization of a landscape; these serai stages terminated in a climax community 
specific to the system. He believed the climax community was an expression of the 
system, and not driven by changes caused by individual species. Disruption of the 
successional sequence by disturbance returned the entire species group, or sere, to an 
earlier serai stage, and the sequence of change would repeat. This idea was challenged 
by Gleason (1926) who recognized that a succession of species may occupy a site, but 
questioned that the sequence and association were predetermined by the system. He 



/ 



18 
believed that the expression of species at a site reflected the available propagule pool and 
variations of the environment, and could be altered by the species present, so that 
seemingly similar sites could be occupied by different individuals, species, and 
associations of species. Change in the composition of the site might occur, but he 
believed that it was not necessarily by predetermined associations of species; the 
response was of the individual, not the association. Gleason believed disturbance 
prohibited a true climax community from developing. Perhaps there is an acceptable 
compromise between these approaches. Succession is a phenomenon of the individual 
and species, not the community, and results from differential life histories, adaptations 
along environmental gradients, and competition among species. Change in an 
individual's environment that exceeds its tolerances may lead to occupation by other 
individuals and species. These changes operate at multiple spatial and temporal scales 
that may be complementary or independent. The selection of species that may occur is 
limited by adaptations to the environment; this gives the appearance of an association of 
species in a community type, but it is on individuals, not the group, that the environment 
exerts control. 

Although disturbance might appear to disrupt an ecosystem, it can also be 
considered a driver of the succession continuum. Disturbance usually adjusts conditions 
to those earlier in the continuum; a cycle of disturbance, occupation, modification, 
development, and repeated disturbance develops. A system's response to the 
disturbance depends on the severity of the disruption and the degree of system 
complexity and development (Holling 1987, Allen and Wyleto 1983). Systems respond 



19 
to disturbance with a period of release and reorganization; a longer state of 
disequilibrium follows disturbance of later succession communities before 
reorganization occurs because more older systems may be less resilient to disturbance 
(Holling 1987). However, the succession sequence repeats in response to the 
disturbance, unless the system has been unnaturally altered. Disruptions in components 
and functions that the altered system experiences may prevent it from developing the 
same hierarchical structure; response of the system to disturbances may then be 
unpredictable and lead to an alternative stability domain (Holling 1987, Forman and 
Godron 1986). 

Monitoring Landscape Change 

Changes in plant communities as they occur within the landscape can be 
monitored using remote sensing and geographical information systems (GIS). 
Frequently, remote sensing provides historic data unavailable in any other format. 
Remote sensing provides data at various temporal and spatial scales at a cost lower than 
required by traditional field censusing techniques, which are used to validate 
interpretations of remotely sensed data with information at greater resolution. These data 
can be combined with other site features (e.g., water chemistry, hydrologic parameters, 
topography, soil type) in a spatially referenced database. Estimation of missing data with 
interpolation may be necessary to provide complete spatial coverage, and data scale must 
be comparable among variables. Cartographic modeling techniques can be used with 



20 
these data to describe relationships among parameters and changes occurring, describe 
how landscape structure influences responses to perturbations, manipulate landscape 
features, and predict spatial effects of these manipulations at various scales (Turner and 
Gardner 1991). Limitations of the data and GIS techniques must be recognized, 
however, so that the influence of data and model scale on interpretations of results is 
understood (Haines- Young and Chopping 1996, Meentemyer and Box 1987). 

The Okefenokee Swamp Ecosystem 

The Okefenokee Swamp is a complex of forested uplands and freshwater 
wetlands covering approximately 1670 km 2 of lower Atlantic Coastal Plain in Ware, 
Clinch, Charlton, and Echols Counties, Georgia, and Baker County, Florida (Figure 1-4). 
Approximately 80% of the swamp is within the Okefenokee Swamp National Wildlife 
Refuge. The geologic origin of the swamp is debated; the traditional theory is that the 
swamp basin began to form during the Yarmouth Interglacial (200,000 years ago) when a 
coastal lagoon became separated from the Atlantic Ocean by a sand bar, known today as 
Trail Ridge (Carver et al. 1986, Cohen 1973b). During the thousands of years following 
this isolation, the seawater evaporated and organism remains and salts were removed by 
water and wind. Climatic changes occurring during the last glaciation brought increased 
precipitation, which collected in the lagoon basin and provided an environment suitable 
for freshwater wetland plants. Peat began to accumulate 6,500 years ago, as decay of 
plant remains was delayed by continuous flooding which created anaerobic, acidic 



21 




Savannah 



♦ Tallahassee . u ♦» Jacksonville 



Okefenokee Swamp 
National Wildlife Refuge 



Figure 1-4. Location of the Okefenokee Swamp and Okefenokee Swamp National 
Wildlife Refuge. 



22 
conditions (Cohen 1973b). Peat accumulation continues today and is punctuated by 
periods of extreme drought, when peat is removed by fire and oxidation. An alternative 
theory initiates basin formation approximately 12,000 years ago; wind scoured the area, 
creating a depression where intercepted rainfall and surface runoff accumulated, and 
decreasing hydraulic head and outflow velocity increased retention time of standing 
water (Parrish 1971). Wetland plants eventually invaded, and the basin filled with 
accumulating peat (Rykiel and Parrish 1979, Rykiel 1977, Parrish 1971). 

The swamp's watershed (3702 km 2 ) includes 3 drainage basins (Brook and Hyatt 
1985, Hyatt 1984, Rykiel 1977). The Suwannee River carries 85% of the exiting flow 
from the western swamp; the St. Marys River (11%) and Cypress and Sweetwater Creeks 
(4%) account for the remainder exiting the southern third of the swamp (Rykiel 1977). 
Groundwater exchange is minimal (Brook and Hyatt 1985, Hyatt and Brook 1984). The 
Suwannee River sill was constructed in 1960 to intercept part of the Suwannee River 
discharge from the swamp; the low, earthen dam was intended to impound water in the 
swamp to protect it from drought, and to control the initiation and spread of wildfires 
within and beyond refuge borders (Chapter 742, Public Law 81-810, 70 Statute 668). 
Discharge from the swamp via the Suwannee River and variability of flow into the St. 
Marys River decreased during 1960-1986, following construction of the sill, whereas 
flow into the St. Marys River increased (Yin and Brook 1992b, Yin 1990). Water enters 
the swamp as precipitation (70%) and surface drainage of uplands along the western and 
eastern boundaries (Rykiel 1977). Water levels are generally lowest during April-May, 
when evapotranspiration demands are high and seasonal precipitation is low, and 



23 
October-November due to low precipitation (Chapter 2). Most rainfall occurs during 

June-September. During periods of normal hydrology, when peat is continuously 

saturated, swamp water depths average 0.7 m (Finn and Rykiel 1979). During the 25 

years following construction of the Suwannee River sill, water depths levels during 

droughts were estimated to be 1 1 cm higher than during pre-sill droughts (Yin nd Brook 

1992b, Yin 1990, Finn and Rykiel 1979). 

Several vegetation communities occur in the Okefenokee Swamp. Prairies are 

found where peat layers are thick over depressions in the basement topography (Cohen et 

al. 1984, Cohen 1974, 1973a, 1973b) and cover approximately 8% of the swamp. 

Vegetation communities include shallow emergent prairies of yellow-eyed grass (Xyris 

spp.) and Walter's sedge {Car ex waiter land) and deeper rooted or floating aquatic 

macrophytes (fragrant water lily, Nymphaea odorata, and golden club, Orontium 

aquaticum). Forested areas of pond cypress {Taxodium acsendens), titi (Cyril la 

racemiflora), hurrahbush (Lyonia lucida), loblolly bay (Gordonia lasianthus), and 

dahoon holly (Ilex cassine) cover 57% of the swamp. Forested uplands of slash pine 

(Pinus elliottii), longleaf pine (P. palustris), saw-palmetto (Serenoa repens), and 

gallberry (Ilex glabra) occur on the remaining area of sandy islands and ridges (5%). 

Dense shrub thickets of titi, hurrahbush, and fetterbush (Leucothoe racemosa), covered 

with a blanket of bamboo greenbriar (Smilax laurifolia) and Walter's greenbriar (S. 

waiter i) fill the remaining swamp interior (29%). Much of the western portion of the 

swamp, where mixed forests of pond cypress, loblolly bay, and blackgum (Nyssa 

sylvatica v. biflora) historically predominated, was logged during 1900-1930 (Mar 



24 
1984). This area currently supports stands of shrubs and hardwoods, with little cypress 
regeneration (Hamilton 1984, 1982). 

The classic model of hydrarch succession (development of a terrestrial forest 
climax community from an open water body) directed by autogenic processes (Mitsch 
and Gosselink 1986) is only partially applicable to the swamp. The topography 
facilitates collection of surface water in the swamp, and periodic droughts expose the 
accumulated peat, allowing oxidation and decline in the surface elevation. Site 
elevations are raised and hydroperiods altered when accumulated peat is not periodically 
exposed and oxidized, creating more favorable conditions for species less tolerant of 
flooding. However, in the swamp's history this exchange of species and apparent 
"progression" have frequently been disrupted when drought, fire, and subsequent species 
changes occur, and the wetland landscape mosaic is maintained (Hopkins 1947). 

Palynological studies suggest that overall plant composition has been similar to 
the current species composition since peat layers began to accumulate (Cohen et al. 1984, 
Rich 1984a, 1984b, 1979, Cohen 1975, 1974, 1973a, 1973b). However, spatial 
distribution of these communities has varied, as is indicated in peat deposits (Cohen et 
al.l984,Rich 1984a, 1984b, 1979, Cohen 1975, 1974, 1973a, 1973b). Hydrologic 
variations and fire interact to direct succession in the swamp (Roelle and Hamilton 1990, 
Hamilton 1984, 1982, Deuver and Riopelle 1984, Duever 1982, 1979,Rykiel 1977). 
Many species occurring in the swamp are adapted to nutrient-poor, saturated conditions. 
Okefenokee Swamp surface waters contain <1% of the system's nutrients; 59-98% of the 
Ca, Mg, Na, and K are found in the system's standing vegetation, and the remainder is 



25 
encumbered in slowly decomposing peat (Rykiel 1977). Exposure of the peat surface 
during drought hastens peat decomposition and bacterial cycling (Murray and Hodson 
1985), making nutrients more available for use (Schoenberg and Oliver 1988, Bosserman 
1983a, 1983b, Flebbe 1983), as do fires which may accompany extended drought. Most 
of the swamp has developed from open prairie to shrub bog to cypress or bay forest 
during some period of the past 6000 years, with undisturbed intervals varying from 
decades to hundreds of years. Drought, peat accumulation, and battery formation reduce 
the apparent water level, which permits succession of flood-intolerant woody species to 
occur. A progression from prairie to cypress swamp to broadleaved evergreen or mixed 
cypress swamp occurs in the absence of disturbance as peat accumulates (Hamilton 1984, 
1982) (Figure 1-5). 

As indicated by layers of charcoal in peat deposits, fire has checked succession in 
the swamp since peat began to accumulate thousands of years ago (Cohen et al. 1984, 
Cohen 1975, 1974, 1973a). Fire retards the progression of prairie to wooded swamp or 
returns the vegetation to and earlier stage. Certain vegetation communities such as 
cypress are frequently associated with concentrations of charcoal in the peat, suggesting 
a susceptibility to fire, especially during droughts (Cohen et al. 1984, Cohen 1975, 1974, 
1973a). The central, deep-peat prairies have never been completely succeeded to cypress 
forest, possibly because they are topographic lows that have maintained conditions too 
saturated for forest species, or severe fire has burned the area frequently enough to retard 
expansion of woody species. Fires which burn the surface peat remove fire-intolerant 
plants but usually do not kill shrubs and large trees rooted in deep peat or sand beneath 



26 






c 

a 

O 



• *■« 

— 

"S3 

o ■ 

cd 

cr 
< 



*C 

2 

Oh 

3 

. O. 

<L> 

o 

cd 




£ 


c 
5 


§ 


■*-» 


o 
u 


E 


o 


C3 


■*-» 


F 


a> 


o 



• T3 
•O (8 



E 

E on 



g 



27 
shallow peat (Cypert 1973, 1961). More severe fires that burn to the deep, sub-peat sand 
layer are disjunct and rarely occur. Deep lakes occurring in the eastern swamp may have 
resulted from hot fires that burned through the accumulated peat and into the underlying 
sand. Prairies result when severe fires remove peat and woody root systems, preventing 
reestablishment of existing woody vegetation (Cypert 1973, 1961) due to lowered 
topographic surface and increased inundation depth. The light, surface fires which 
historically occurred frequently are probably more important than the infrequent, 
widespread, severe fires in maintaining the mosaic of existing vegetation associations 
(Roelle and Hamilton 1990). The manipulations of the swamp vegetation composition 
and hydrology during the past two centuries and current fire management have affected 
fire frequency and occurrence across the swamp (see Human Modification of 
Okefenokee Swamp section). 

The Okefenokee Swamp landscape structure is affected by vegetation community 
succession. Swamp vegetation is determined by the hydrologic environment; disturbance 
history; species pool; propagule distribution and establishment requirements; potential 
longevity of propagules, juveniles, and adults; and, interactions of these features. 
Although species composition and abundance vary from site to site, there is a limited 
number of species occurring in the swamp, each with a certain range of life history 
requirements and environmental tolerances. Thus only a certain suite of species are 
likely to occur, and their presence in the landscape is mediated by inter- and intraspecific 
interactions, as well as other environmental processes. The swamp is maintained as a 
metastable equilibrium, where species are fluctuating between a competitive equilibrium 






28 
(maintaining the appearance of stability) and disequilibrium (species replacement), with 
intervening periods of changing communities in response to disturbance of greater 
intensity on a larger spatial scale. In the swamp's pre-modern (pre- 1 890) history this 
disturbance has been periods of drought and fire, creating a moving mosaic of vegetation 
communities in different stages of development in the landscape (Hamilton 1984, 1982, 
Rykiel 1977). Hydrologic alterations, logging, and changes in the burning regime are 
perturbations with the potential to disrupt development of this mosaic and affect future 
swamp structure. 

Human Modification of Okefenokee Swamp 

Humans have inhabited the Okefenokee Swamp for at least the past 4,000 years, 
and have lived in the Okefenokee Swamp area for 10-12,000 years (Trowell 1984a, 
1984b), and have variously modified the swamp, particularly during the 20th century 
(Trowell 1994, 1989a, 1989b, 1988a, 1988b, 1987, 1984) (Table 1-1). The region's 
name is derived from a Seminole-Creek Nation word, Oke-fm-o-cau, meaning "land of 
the trembling earth" (McQueen and Mizell 1926), in reference to the floating islands 
found throughout the swamp. The swamp was surveyed by Mansfield Torrance in 1850 
and many others in the following years, and was purchased from the State of Georgia by 
the Suwannee Canal Company in 1890 (Trowell 1994, 1989a, 1989b, 1988a, 1988b, 
1984a, 1984b). The Suwannee Canal was excavated during 1890-1897 to drain the 
swamp to create an agricultural district; the effort failed, and in 1904 the land was 



29 



Table 1-1. Human-caused manipulations of Okefenokee Swamp vegetation and 
topography occurring during the past 1 50 years. 



Type of 
Manipulation 


When Manipulation 
Occurred 


Probable Scale 
of Effect 


Prescribed Burning, Arson 


pre-settlement to present 


local to swamp-wide 


Dredging, Peat Mining 


late 19th century to mid- 
20th century 


local peat removal, 
regional hydrologic effect 


Logging 


late 19th century to mid- 
20th century 


locally intensive, 
regionally scattered 


Impoundment 
(Suwannee River sill) 


1 960- 1 962 construction, 
1962 fully operational 


local to regional effects 
depending on seasonal 
water levels 


Boat Trail Cutting and 
Maintenance 


20th century 


local to regional effects on 
submerged, emergent, and 
nearby terrestrial 
vegetation 



purchased by the Hebard Lumber Company. Marketable cypress, pine, and hardwoods 
were removed from the swamp and processed at local sawmills for shipment throughout 
the Southeast. Logging operations ceased in 1937 when the property was purchased by 
the United States government and added to the National Wildlife Refuge system. Peat 
mining in the Northeast swamp ceased in the 1950s, and the refuge was designated a 
national wilderness area in 1974 (Trowell 1989c, Fortson 1961). 

During 1954-1955 the region experienced a severe drought and nearly 80% of the 
swamp was burned by wildfires (Hamilton 1984, 1982). Many of these fires began in the 
surrounding uplands, spread into the swamp where the peat slowly burned, and returned 
to the perimeter uplands. Neighboring landowners sustained significant property loss 



30 

from these fires. There was great interest in protecting the swamp and surrounding lands 

from future fires; a law (Appendix A) was enacted by the United States Congress in 1956 

to require construction of a dam, the Suwannee River sill, 

to protect the natural features and the very substantial 
public values represented in the Okefenokee National 
Wildlife Refuge, Georgia, from disastrous fires..., and for 
the purpose of safeguarding the forest resources on more 
than four hundred thousand acres of adjoining lands 
recently damaged by wildfires originating in or sustained 
by the desiccated peat deposits in the Okefenokee Swamp. 
(Chapter 742, Public Law 81-810, 70 Statute 668, pages 
781-782). 

A perimeter road that would permit access to remote areas for fire control and serve as a 

fire break to spreading fires was also required by the law. In 1962 construction of the sill 

berm and closure of the 2 spillway gates were completed. The berm spans 7.2 km across 

the exiting flow of the Suwannee River and averages 35.5m above mean sea level and 3- 

4 m above the surrounding Suwannee River floodplain; a ditch borders its entire length 

to the east. The original south gate collapsed in 1979 and was replaced; the north gate is 

the original structure. Although the gates are maneuverable, they remain closed to 

maximize the impoundment. 

Apparent changes in vegetation composition of the Okefenokee Swamp during 

1960-1990 precipitated concern that the Suwannee River sill and the Okefenokee 

National Wildlife Refuge fire management policy were permanently altering the 

swamp's ecology (Roelle and Hamilton 1990). Severe drought and fire have occurred in 

the Okefenokee Swamp at approximately 20-year intervals during the past 150 years 

(Cypert 1973, 1961). The Suwannee River sill was constructed to prevent recurrence of 



T 



31 
fires during these droughts. During 1962-1990 extensive fires did not occur in the 

swamp. This may have been the result of the Refuge's fire management policy rather 

than the impoundment effects of the sill. Yin and Brook (1992b) and Yin (1990) found 

that the amount of water retained by the sill during severe drought (1 1 cm) was not 

enough to counteract an extreme drawdown (1-1.5 m during 1954-1955) due to drought. 

In fact, scattered fires during 1990 and 1993 suggest that the sill had not eliminated fire 

in that region. Thus, the sill was performing as it was intended (i.e.,to suppress fires) 

only in its localized area during periods of average hydrologic conditions, temporally and 

spatially extending hydroperiod beyond the local area during intervening years when 

water levels were generally higher (Roelle and Hamilton 1990), and not retaining a 

substantial amount of water during extended periods of below average rainfall. 

Extending flooding by impounding runoff and stream flow may reduce water 

level variation that normally occurs with precipitation (Finn and Rykiel 1979). Finn and 

Rykiel (1979) compared pre- and with-sill water levels measured at the Camp Cornelia 

boat basin 29 km east of the sill, and reported an increase (10-13 cm) in average monthly 

water level after sill construction. Yin and Brook (1992b) and Yin (1990) measured an 

increase in average storage and a decrease in discharge. The higher water level behind 

the sill decreases the gradient approaching the sill, reducing flow and pooling the water 

(Finn and Rykiel 1979), especially during periods of above average rainfall. If the sill is 

extending periods of high water, it may be altering the landscape by affecting vegetation 

succession. Decreased fire frequency and extent may be encouraging woody vegetation 

to invade prairies during the occasional drier periods, hastening succession to cypress or 



32 
bay swamp, and eliminating the mosaic of vegetation and the associated biodiversity in a 
landscape historically perpetuated by periodic disturbances (Hamilton 1984, 1982). 

When this study was initiated in late 1991, the Suwannee River sill had 
deteriorated since its construction and was in need of repair. The uncertainty of the sill's 
effects on the hydrology and vegetation of the swamp raised questions of whether the sill 
should be opened, repaired as a fixed height weir, or replaced with a controllable 
structure. Effects of the sill on vegetation communities within the landscape needed to 
be documented and predicted effects of future hydrologic management alternatives 
analyzed so that the refuge hydrology could be effectively managed. This dissertation 
research identifies the spatial extent of the Suwannee River sill's modification of swamp 
hydrology, and spatial changes in vegetation composition since the sill was constructed. 
Probable causes of the vegetation changes are proposed, and several hydrology 
management options and their effects on swamp vegetation composition are investigated. 
The following guiding questions are addressed in these dissertation chapters: 

1 ) Have vegetation community distributions changed since 
the Suwannee River sill was constructed? If so, where have 
these changes occurred? Have fire frequency and distribution 
changed during this period? 

2) Are swamp vegetation community composition and distribution 
correlated with hydroperiod and water depth? 



33 
3) What are the potential responses of the Okefenokee Swamp 

vegetation communities and landscape to future sill modification 
and hydrologic manipulation? 



CHAPTER 2 
DATA BASE ORIGIN AND DEVELOPMENT 



Data Sources and Extent 



Determining the effects of the Suwannee River Sill on the hydrology and 
vegetation of Okefenokee Swamp required diverse point and spatial data. The origin, 
management, and quality assessments of data used in the swamp hydrology model (see 
Chapter 3) and vegetation change analysis (see Chapter 4) are detailed in this chapter. 

Precipitation, evapotranspiration, surface water inflow and outflow, water surface 
elevation, and water depth data are components of the swamp hydrology model. The 
hydrology model describes the swamp surface water environment during 1941-1993, the 
duration of the complete, concurrent weather and water level data. Summaries for 
Suwannee Canal Recreation Area (SCRA) and Stephen C. Foster State Park (SCFSP) 
include the complete period of record for these stations, 1941-1995. Historic, daily data 
were available for 1930-1991 from gauges monitoring some of the model parameters; 
additional data were collected during 1991-1995 from gauges installed in 1991-1992 to 
supplement the recorder network. Descriptive statistics for precipitation, 



34 



35 
evapotranspiration, and surface water flow are calculated from the 1930-1995 data. Due 
to recorder discontinuity, malfunction, or removal, the daily records for these parameters 
during 1930-1995 were incomplete. Regression equations between correlated recorders 
estimated missing data to provide a more complete data record for the hydrology model. 
Descriptions of hydrology dataset management and the swamp hydrologic environment 
are discussed in the Swamp Water Level Data, Estimation of Missing Water Level Data, 
General Swamp Water Level Conditions, and the Suwannee River Sill's Effects on River 
Flow and SCRA and SCFSP Water Level Conditions sections of this chapter. 

Swamp water level variation was monitored with daily data from a water level 
recorder network. Network design redundancies and discrepancies affect the accuracy of 
the water level estimates. Identifying the best design, to improve efficiency and accuracy 
of the recorder network for the intended purpose, is discussed in the Water Level 
Recorder Performance section of this chapter. 

Most of the water in the swamp enters as direct precipitation (Yin and Brook 
1992b, Yin 1990, Hyatt 1984, Blood 1981, Rykiel 1977). Estimations of spatial 
contributions of rainfall to the swamp water budget rely on data gathered at precipitation 
recorder stations. Accuracy of the recording network should be quantified so that the 
accuracy and limitations of the precipitation estimates can be identified. A rainfall 
variation and recorder network design analysis are discussed in the Precipitation Gauge 
Network Analysis section of this chapter. 

Swamp topographic surface elevation is a component of the swamp hydrology 
model. Although 1994, 1:24,000 USGS topographic maps exist for the swamp region, 



36 

the data scale is insufficient for directing water movement across the slight gradient of 
the swamp at the hydrology model scale (500 x 500 m cell size). A more resolute 
topographic map was developed with a Global Positioning Systems (GPS) survey; this 
survey also permitted referencing the network of recorders to a common reference 
(elevation above mean sea level) and identifying their true location in the landscape 
within centimeters. The data collection and interpolation procedures used to create the 
swamp topographic surface map, and the development of swamp peat and sand surface 
profiles, are discussed in the Topography Map Development Section of this chapter. 

A base map of current vegetation was needed to identify changes in vegetation 
community composition and distribution occurring in the landscape since the sill was 
constructed. SPOT multispectral and panchromatic satellite imagery were used to 
produce this base map; changes in swamp vegetation community distributions were 
identified by comparing maps created from interpretations of aerial photographs of the 
swamp taken in 1952 (7 years pre-sill) to those from 1977 (15 years with-sill) and the 
1990 (28 years with-sill) base map. The procedures and accuracy of the satellite imagery 
classification are detailed in the Satellite Imagery Classification and Accuracy 
Assessment section of this chapter. Details of the photointerpretation and change 
assessment procedures are included in Chapter 4. 
Swamp Water Level Data 

Swamp water level data were compiled from several sources. The longest 
duration records were from staff gauges installed in Billy's Lake (at SCFSP) and the 



37 
SCRA boat basin during 1941; readings were made several times monthly at both 
stations until 1950, when daily readings were begun at SCRA. Daily readings were not 
made at SCFSP until 1968. Steven's chart recorders with float gauges were installed at 
SCRA in 1979 and SCFSP in 1980. Chart recorders were also installed at 1 1 other sites 
in 1979-1980 (Figure 2-1). Elevations of these recorders were referenced to staff gauges 
installed on site, and the reference elevation was transit-surveyed to perimeter USGS 
benchmarks during the early 1980s. During 1980-1991, 3 recorders were removed from 
the network, and those that remained were not regularly maintained, resulting in an 
incomplete record of daily water surface elevation. In 1992 the gauging network was 
examined, broken gauges were repaired, and an additional 13 gauges (Omnidata, Inc. 
digital recorders with Delta pressure transducers and WaterMark reference staff gauges) 
were installed throughout the swamp (Figure 2-1). Elevation of the reference staff at 
each recording station was related to a permanently established benchmark located 
within 500 m of the recorder. Location and elevation of these benchmarks were surveyed 
(see discussion of topography map development in this chapter) relative to Universal 
Transverse Mercator (UTM) zone 17 grid X and Y location and NAVD27 elevation 
projection of mean sea level, so that water surface measurements could be compared 
spatially. The digital gauges recorded water elevation once daily (hourly readings 
averaged every 24 hours); data were retrieved from the recorders every 4-6 weeks during 
1992-1995. Daily water surface elevation recorded on the accumulated historic charts 
(1980-1991) and those retrieved quarterly from chart recorders during 1992-1995 were 
digitized and corrected to the reference benchmark elevation. Records from each station 



1 



38 






fSI 



= ^D WTO 



(3 



CO 

ft 

8 



c75 



*» < < | 







as 

c 

1 

a. 




I 
J 

8 



1 

§ 

I 

— 



5 



1 



<N 



I 



39 

were compiled into spreadsheets; intervals of missing data were identified and 

correlations among recorders examined to identify regression equations to use in missing 
data estimation (see Recorder Correlations and Missing Data Estimation). 
Water Level Recorder Performance 

There were 26 gauges recording water elevation continuously or daily during 
1979-1995 for varying lengths of time. Elevation of reference staffs above mean sea 
level, corrections to historic reference staff data, period of record, and days in operation 
are indicated in Table 2-1 . The interpreted data from each recorder and staff during the 
periods of operation are illustrated in Figure 2-2. Estimated missing data are included in 
the plots to approximate a complete record for 1941-1995. Water level data were 
estimated for all recorder stations from SCRA and SCFSP staff data during 1941-1979. 
Estimates were calculated for 1-80% of the station water level data for 1980-1995; 
descriptive statistics of each station's record are listed in Table 2-2. 

During 1992-1995 when the water level recorder network density was highest, 23 
gauges were working for 20-99% of the interval (Table 2-3). In most cases recorder 
malfunction could be attributed to mechanical failure due to interference by wildlife or 
refuge visitors, or due to insufficient maintenance of recording equipment. During 1992- 
1995 digital recorders were most reliable, although one chart recorder operated for 91% 
of the interval. The poorer performance of chart recorders can be attributed to their age. 
Most of the stations had been deployed since early 1980. Solar rechargeable batteries 
caused problems with 3 digital units, and insufficient charges on non-rechargeable 
batteries were responsible for missing data on other units. Over the 694-5672 days of 



40 



Table 2-1 . Water level recorder elevations and staff corrections, operating period, and 
precipitation gauge locations. 



Station 


Type* 


Start Date 


End Date 


Staff 

Correction 

(m) 


Ground 
Elevation 

Above 

Mean Sea 

Level at 

Staff (m) 


Sill (Brown 
Trail) 


chart, wl, p 


2-1-1980 


6-7-1995 


+34.75 


34.03 


Chase Prairie 


chart, wl, p 


5-6-1980 


4-10-1995 


+32.24 


36.21 


Territory Prairie 


chart, wl, p 


5-6-1980 


3-13-1995 


+30.32 


36.66 


SCFSP 


chart, wl, p 


2-1-1980 


2-22-1995 


-0.009 


34.12 


SCRA 


chart, wl, p 


9-1-1979 


7-3-1995 


+36.14 


36.08 


Double Lakes 


chart, wl, p 


5-21-1980 


2-22-1995 


+36.18 


37.35 


Gannett Lake 


chart, wl, p 


6-4-1980 


1-19-1995 


-0.63 


36.53 


Seagrove Lake 


chart, wl, p 


12-5-1979 


2-16-1995 


+31.11 


36.50 


Moonshine 
Ridge 


chart, wl, p 


3-5-1982 


5-16-1994 


-1.13 


35.62 


Suwannee Creek 


chart, wl, p 


5-22-1980 


10-17-1982 


+35.97 


36.02 


Soldier's Camp 


chart, wl, p 


4-11-1980 


3-5-1982 


+33.83 


34.0 


Sapp Prairie 


chart, wl, p 


4-18-1980 


3-9-1988 


+35.97 


35.20 


Kingfisher 
Landing 


chart, wl, p 


1-11-1980 


6-14-1995 


+37.38 


36.54 


Coffee Bay 


digital, wl, p 


4-1-1992 


6-14-1995 


+35.77 


36.69 


Billy's Lake 


digital, wl 


4-16-1992 


6-13-1995 


+33.58 


33.6 


Suwannee River 


digital, wl 


5-7-1992 


5-31-1995 


+33.77 


33.62 


Sweetwater 
Creek 


digital, wl 


4-2-1992 


6-1-1995 


+33.84 


34.04 


Cypress Creek 


digital, wl 


4-3-1992 


6-1-1995 


+32.94 


33.95 


Floyd's Prairie 


digital, wl, p 


4-2-1992 


6-1-1995 


+34.86 


35.27 


Suwannee Creek 


digital, wl 


4-17-1992 


7-20-1995 


+36.66 


36.02 



Table 2-1 -continued. 



41 













Ground 












Elevation 










Staff 


Above 


Station 


Type' 


Start Date 


End Date 


Correction 
(m) 


Mean Sea 
Level at 
Staff (m) 


Sapling Prairie 


digital, wl, p 


2-4-1993 


2-22-1995 


+36.35 


36.61 


Durdin Prairie 


digital, wl, p 


4-1-1992 


6-14-1995 


+36.68 


36.98 


Honey Prairie 


digital, wl, p 


6-16-1992 


12-15-1994 


+35.84 


36.11 


Chesser Prairie 


digital, wl 


2-4-1993 


6-14-1995 


+35.94 


35.68 


Sapp Prairie 


digital, wl, p 


2-4-1993 


12-15-1994 


+35.70 


35.20 


Craven's 


digital, wl, p 


4-1-1992 


5-31-1995 


+34.95 


35.15 


Hammock 












SCFSP 


staff, wl 


1-4-1941 


2-22-1995 


-0.009 


34.12 


SCPvA 


staff, wl 


1-4-1941 


7-3-1995 


+36.14 


36.08 



* Data are recorded daily by automated (chart, digital) systems or refuge personnel (staff), and 
stations monitor daily water surface elevations (wl), precipitation (p), or both. 



42 



CO 



c 
g 

CO 

> 

LU 



37 
36 



33 



37 



CO 

< 

c 
o 
■-•-' 

> 

111 



36 
36 - 
34 - 



33 



37 



CO 



c 
o 
"■«-' 

(0 

> 

Hi 



36 - 



35 



34 - 



33 



Billy's Lake 




i i i i 1 1 1 — 

1941 1944 1947 1950 1953 1956 1959 




Uv\a/ 



1960 1963 1966 1969 1972 1975 1978 




1980 1983 1986 1989 1992 1995 

Year 



Figure 2-2. Daily water surface elevation above mean sea level (AMSL) during 1941 -May 
1995 recorded at locations in Okefenokee National Wildlife Refuge, GA. 



43 



37.0 



CO 36.5 




1941 1944 1947 1950 1953 1956 1959 



37.0 



35.5 



37.0 




1 1 I l 1 1 1 — 

1960 1963 1966 1969 1972 1975 1978 



CO 36.5 - 




Year 



Figure 2-2 - continued. 



CO 



c 
o 
'■*-> 

(0 

> 

LLI 




35.0 



37.5 



1941 



«- 37.0 - 

CO 

2 36.5 - 



c 
g 

> 

LLI 



36.0 - 



35.5 - 



35.0 



37.5 



1960 



CO 



g 

(0 

> 

LLI 



1980 



1944 



1947 



1950 1953 



1956 



1959 




1963 



1966 



1969 



1972 



1975 



1978 




1983 



1986 1989 

Year 



1992 



1995 



Figure 2-2- continued 



38 



35 
38 



CO 37 - 



c 

■I 36 - 

> 



LU 



35 



38 



E 

CO 37 - 



c 

H 36 

> 

LU 



35 



Coffee Bay 




T 



T 



1941 1944 1947 1950 1953 1956 



T 



1980 1983 1986 1989 

Year 



1992 



1959 



T 



1960 1963 1966 1969 1972 1975 1978 




45 




1995 



Figure 2-2- continued 



CO 



37 



36 - 



§ 35- 

CO 

> 
iH 34 



1941 



37 



CO 



36 - 



c 
o 


35 


-♦— * 




CO 




> 





HI 


34 



1960 



37 



CO 



36 - 



| 35 

'■*-> 

CD 

> 

UJ 34 



46 



Craven's Hammock 




1944 1947 1950 1953 1956 1959 




1963 1966 1969 1972 1975 1978 




1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2 -continued 



47 



E 35- 



co 



c 
o 



34 - 



ro 33 - 

HI 



32 



Cypress Creek 




1941 1944 1947 1950 1953 1956 1959 



E 35- 



34 - 



CO 



c 
g 

ro 33 

111 
32 




1 I 

1960 1963 



I I I 1 1 — 

1966 1969 1972 1975 1978 



E 35- 



CO 



c 
o 



34 



ro 33 - 

0) 

LU 



32 




1980 1983 1986 1989 

Year 



1992 1995 



Figure 2-2-continued 



co 



c 
g 

"co 

> 

HI 



38.0 



37.8 - 



37.6 



37.4 - 



37.2 - 



37.0 



38.0 



CO 

< 

c 
o 

(0 

> 

LU 



37.0 



38.0 



CO 



c 
g 

> 

0) 
LU 



48 




i r 

1941 1944 1947 1950 



1953 1956 



1959 




1960 1963 1966 1969 1972 1975 1978 




1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2-- continued. 






49 



37.6 




37.0 
37.6 



1960 



37.0 



1963 



1966 



1969 



1972 



1975 



1978 




I I 1 1 — 

1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2- continued 






CO 



C 

o 

CO 

> 

LU 




35.0 - 



1941 



37.0 



CO 



c 
o 

CD 

> 
0) 

LU 



36.5 - 



36.0 - 



35.5 - 



35.0 - 



34.5 



37.0 



1960 



CO 



g 

> 
LU 



1944 



1947 



1950 



1953 



1956 1959 




1963 



1966 1969 



1972 1975 



1978 




1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2--continued 



51 



37.5 



<f) 37.0 - 




1941 



37.5 



37.5 



CO 37.0 - 



1944 1947 1950 1953 1956 1959 




1960 1963 1966 1969 1972 1975 1978 




Year 



Figure 2-2--continued. 



38.0 



38.0 



_i 37.5 - 

CO 



c 
o 



37.0 - 



2 36.5- 

LU 



36.0 



LU 



36.0 



52 



Honey Prairie 




IT 37.5 - 

CO 


/\.«Jl A A* FVt 


< 37.0 - 

c 
g 

1 36.5- 



tUvY^ 



T 



1980 1983 1986 1989 

Year 



1992 



1941 1944 1947 1950 1953 1956 1959 




1960 1963 1966 1969 1972 1975 1978 



1995 



Figure 2-2-continued 



53 



38.0 



CO 

< 37.5 - 

c 
g 
■*- 

(0 

> 

LU 37.0 - 



Kingfisher Landing 




1 - ! ! j ! ! ! — 

1941 1944 1947 1950 1953 1956 1959 



38.0 



CO 

< 37.5 - 

c 
o 

«SS 
03 

W 37.0 - 




1^ I I I | I 1 — 

1960 1963 1966 1969 1972 1975 1978 



38.0 



E 

-J 
CO 

< 37.5 - 

c 
g 
*-• 

(0 

> 

LU 37.0 - 




i r 



1980 1983 



1986 1989 

Year 



1992 1995 



Figure 2-2- continued 



54 



CO 



c 
o 



36.0 



2 35.5 

LU 




1941 



i i 1 r 

1944 1947 1950 1953 1956 1959 



CO 



c 
o 



36.0 - 



2 35.5 
LU 




1960 



i i i i r 

1963 1966 1969 1972 1975 1978 



_j 36.0 - 
CO 



c 
g 

I 35.5 


LU 




1980 1983 1986 1989 1992 1995 

Year 



Figure 2-2- continued 



55 



CO 



c 
g 

CO 

> 
111 



38.0 
37.5 



36.0 



38.0 



T 



_j 37.5 - 

CO 



c 
g 

CO 

> 

LU 



37.0 - 

36.5 

36.0 



38.0 



CO 



c 
q 

(0 

> 

LU 



37.5 - 



37.0 



36.5 - 



36.0 



Sapling Prairie 




^^ i i i 1 1 — 

1941 1944 1947 1950 1953 1956 1959 




1980 1983 1986 1989 

Year 



1960 1963 1966 1969 1972 1975 1978 




1992 



1995 



Figure 2-2 -continued. 



56 



E, 37.0 - 

_j 
CO 

I 36.5 - 

c 
o 

ro 36.0 - 

LU 



35.6 



1941 



35.5 



Sapp Prairie 




i 1 1 1 r 

1944 1947 1950 1953 1956 1959 




1 I 1 1 1 1 1 — 

1960 1963 1966 1969 1972 1975 1978 




1980 1983 



1986 1989 

Year 



1992 1995 



Figure 2-2- continued. 



37 



— 1 


36 - 


CO 




2 




< 


35 - 


c 




o 




*-> 




(0 

> 


34 - 


<D 





LU 



33 



37 



1941 



—1 


36 


CO 




2 




< 


35 


c 




o 




,»■ ■ 




(0 

> 


34 


0) 





LU 



33 



37 



1960 



-J 


36 


CO 




2 




< 


35 


c 




o 




■*— ' 




(0 

> 


34 


o 





LU 



33 



57 



SCFSP 




1944 



1947 



1950 



1953 



1956 1959 




T 



T 



1963 



1966 



1969 



1972 1975 



1978 




1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2-continued 



58 



CO 



c 
g 

> 
LU 



35.5 



37.5 



CO 



c 
g 

m 

> 

LU 



36.0 - 



35.5 



37.5 



CO 



c 
o 

(0 

> 

0) 
LU 



37.0 - 



SCRA 




i i i i i 

1941 1944 1947 1950 1953 



— I 1 — 

1956 1959 




1 I I I l 
1960 1963 1966 1969 1972 



1980 1983 1986 1989 

Year 



1992 



1975 1978 




1995 



Figure 2-2 -continued. 



CO 



c 
o 

'■*-• 
CO 

> 

LU 



37.5 



37.0 - 



36.5 - 



36.0 



59 



35.5 



37.5 



_j 37.0 - 
CO 



c 
o 

CD 

> 

LU 



36.5 



36.0 - 



35.5 



37.5 



CO 



c 
o 

CD 

> 

LU 



37.0 - 
36.5 - 
36.0 - 
35.5 



1980 



T 



Seagrove Lake 




1941 1944 1947 1950 1953 1956 1959 



1983 



1986 1989 

Year 



1992 




i I I I I — 
1960 1963 1966 1969 1972 1975 1978 




1995 



Figure 2-2-continued 



CO 



c 
o 

'■4-< 

C5 
> 

HI 




1941 1944 1947 1950 1953 1956 1959 



CO 

I 

c 
o 

'■*-> 
CO 

> 

HI 




1960 1963 1966 1969 1972 1975 1978 



CO 

< 

c 
o 

*-» 
CO 

> 

HI 




1980 1983 1986 1989 1992 1995 

Year 



Figure 2-2- continued. 



61 



E 


35.0 - 


"»»—■*• 




-j 




CO 




2> 


34.5 - 


< 




c 




o 


34.0 - 


•+-> 




(D 




> 




<D 




UJ 


33.5 - 



E 35.0 



Soldier's Camp 




-| r 1 r- 

1941 1944 1947 1950 1953 1956 1959 




1960 1963 1966 1969 1972 1975 



1978 




1980 



1983 



l I 
1986 1989 

Year 



1992 



1995 



Figure 2-2- continued 



62 



38.5 



CO 



c 

CO 

> 

0) 

LU 



38.5 



CO 



c 
o 

CO 

> 

LU 



Suwannee Creek 




i i^ n i i i i 

1941 1944 1947 1950 1953 1956 1959 




1960 1963 



38.5 



CO 



g 

"CO 

> 

LU 



1966 



1969 



1972 



1975 



1978 




I I I I 
1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2--continued. 



CO 



c 
o 

> 
111 




i i r 

1941 1944 1947 1950 1953 1956 1959 



CO 



c 
o 

> 
a> 

LU 




1960 



1963 1966 1969 1972 1975 1978 



CO 

< 

c 
o 

'■♦3 
(0 

> 

UJ 




1980 1983 1986 1989 

Year 



1992 



1995 



Figure 2-2 -continued 



64 



£ 35.0 - 

_i 
CO 

I 34.5- 

c 
g 

| 34.0 - 

o 

LU 



33.5 



Sweetwater Creek 




1941 1944 1947 1950 1953 1956 1959 




1960 1963 1966 1969 1972 1975 1978 




1980 1983 



1986 1989 

Year 



1992 1995 



Figure 2-2 -continued. 



65 



37.5 



CO 37.0 - 




1941 1944 1947 1950 1953 1956 1959 



37.5 



CO 37.0 - 




t i r 

1960 1963 1966 1969 1972 1975 1978 



37.5 




1980 1983 1986 1989 1992 1995 

Year 



Figure 2-2- continued 



66 




Table 2-2. Summary parameters of water level recorders installed at Okefenokee 
National Wildlife Refuge during 12-5-1979 through 6-15-1995. Elevations are in 
meters above mean sea level. Basin delineation is discussed in the Swamp Basin 
Delineation and Characterization section. 






Basin and 
Station" 


Mean Daily 

Water Surface 

Elevation 


Variance in 

Daily Water 

Surface 

Elevation 


Minimum 

Daily Water 

Surface 

Elevation 


Maximum 

Daily Water 

Surface 

Elevation 




Northwest 












Basin 














Suwannee 
Creek (digital) 


36.96 


0.12 


36.13 


38.07 






Suwannee 
Creek (chart) 


36.39 


0.04 


35.77 


36.76 






Floyd's Prairie 


35.64 


0.02 


35.26 


36.03 






Sapling Prairie 


37.08 


0.02 


36.74 


37.36 






Suwannee 
River 


34.65 


0.27 


33.31 


35.39 






Billy's Lake 


35.03 


0.05 


34.53 


35.52 






SCFSP 


34.95 


0.08 


34.07 


36.04 






Sill (Brown 
Trail) 


35.08 


0.23 


33.29 


36.10 






Craven's 
Hammock 


35.51 


0.08 


35.05 


36.11 






Northeast 














Basin 














Kingfisher 
Landing 


37.55 


0.02 


37.25 


37.86 






Double Lakes 


37.54 


0.02 


37.16 


37.86 






Durdin Prairie 


37.35 


0.004 


37.17 


37.49 






Central Basin 














SCRA 


36.52 


0.05 


35.72 


37.04 






Seagrove Lake 


36.63 


0.04 


36.00 


37.10 





Table 2-2 --continued. 



67 



Basin and 
Station" 


Mean Daily 

Water Surface 

Elevation 


Variance in 

Daily Water 

Surface 

Elevation 


Minimum 

Daily Water 

Surface 

Elevation 


Maximum 

Daily Water 

Surface 

Elevation 


Chase Prairie 


36.49 


0.01 


36.75 


36.02 


Gannett Lake 


36.78 


0.03 


36.17 


37.17 


Territory 
Prairie 


36.90 


0.03 


36.37 


37.31 


Chesser Prairie 


36.63 


0.02 


36.26 


36.91 


Coffee Bay 


36.81 


0.04 


36.13 


37.08 


Sweetwater 
Creek 


34.56 


0.02 


34.11 


35.07 


Honey Prairie 


37.15 


0.01 


36.86 


37.31 


Southeast 
Basin 










Moonshine 
Ridge 


35.81 


0.01 


35.55 


36.01 


Soldier's Camp 


34.18 


0.05 


33.70 


34.59 


Southwest 
Basin 










Sapp Prairie 
(chart) 


36.61 


0.01 


36.36 


36.91 


Sapp Prairie 
(digital) 


36.47 


0.02 


36.13 


36.72 


Cypress Creek 


34.14 


0.06 


33.35 


34.69 



Operating interval duration for each station is included in Table 2-1. 



68 



Table 2-3. Summary of water level and precipitation recorder performance during 
12-5-1979 through 6-15-1995 at Okefenokee National Wildlife Refuge. 



Station 


Type of Data 
Collected 


Duration of 

Recorder 

Installment 

(days) 


Duration of 

Recorder 

Operation 

(days) 


Proportion of 
Days 

Recorder 
Functioning 

Properly 


Chase Prairie 


water level 


5519 


4358 


79 


Double Lakes 


water level 


5672 


4343 


77 


Gannett Lake 


water level 


5490 


3843 


70 


Kingfisher 
Landing 


water level 


5635 


3553 


63 


Moonshine 
Ridge 


water level 


4851 


2277 


47 


Sapp Prairie 
(chart) 


water level 


2882 


1482 


51 


SCFSP 


water level 


5672 


4071 


72 


SCRA 


water level 


5672 


5017 


88 


Seagrove 
Lake 


water level 


5672 


4328 


76 


Sill (Brown 
Trail) 


water level 


5614 


2649 


47 


Soldier's 
Camp 


water level 


694 


681 


98 


Suwannee 
Creek (chart) 


water level 


878 


854 


97 


Territory 
Prairie 


water level 


5519 


3748 


68 


Billy's Lake 


water level 


1156 


1141 


99 


Chesser 
Prairie 


water level 


862 


791 


92 


Coffee Bay 


water level 


1172 


967 


83 



Table 2-3-continued. 



69 













Station 


Type of Data 
Collected 


Duration of 

Recorder 

Installment 

(days) 


Duration of 

Recorder 

Operation 

(days) 


Proportion of 
Days 

Recorder 
Functioning 

Properly 


Craven's 
Hammock 


water level 


1171 


1039 


89 


Cypress Creek 


water level 


1169 


1036 


89 


Durdin Prairie 


water level 


1102 


963 


87 


Floyd's 
Prairie 


water level 


1170 


1117 


95 


Honey Prairie 


water level 


1101 


224 


20 


Suwannee 
River 


water level 


1170 


1060 


91 


Sapp Prairie 
(digital) 


water level 


862 


677 


79 


Sapling 
Prairie 


water level 


862 


639 


74 


Suwannee 

Creek 

(digital) 


water level 


1155 


975 


84 


Sweetwater 
Creek 


water level 


1170 


883 


75 


Craven's 
Hammock 


precipitation 


1162 


997 


86 


Coffee Bay 


precipitation 


1171 


960 


82 


Durdin Prairie 


precipitation 


1176 


1051 


89 


Floyd's 
Prairie 


precipitation 


1156 


1103 


95 


Honey Prairie 


precipitation 


924 


236 


26 


Sapling 
Prairie 


precipitation 


749 


635 


85 





Table 2-3~continued 






7C 


1 




Station 


Type of Data 
Collected 


Duration of 

Recorder 

Installment 

(days) 


Duration of 

Recorder 

Operation 

(days) 


Proportion of 
Days 

Recorder 
Functioning 

Properly 




Sapp Prairie 
(digital) 


precipitation 


684 


663 


97 




Suwannee 
River 


precipitation 


1126 


1033 


92 






SCFSP 


precipitation 


5501 


2407 


56 






Double Lakes 


precipitation 


5375 


3469 


65 






SCRA 


precipitation 


5784 


4034 


70 






Chase Prairie 


precipitation 


5454 


3650 


67 






Seagrove 
Lake 


precipitation 


5553 


3342 


60 






Kingfisher 
Landing 


precipitation 


5634 


3017 


54 






Gannett Lake 


precipitation 


5344 


3853 


72 






Territory 
Prairie 


precipitation 


5425 


3523 


65 






Sill (Brown 
Trail) 


precipitation 


5605 


2798 


50 






Moonshine 
Ridge 


precipitation 


4458 


2398 


54 






Suwannee 
Creek (chart) 


precipitation 


878 


766 


87 






Soldier's 
Camp 


precipitation 


704 


552 


78 






Sapp Prairie 
(chart) 


precipitation 


2888 


1161 


40 



















71 

chart recorder operation, 46% operated for >75% of the installation period. Over the 
installation period of digital recorders (862-1 172 days), 85% functioned for >75% of 
the interval. If the operation period is pro-rated to the same length for both recorder 
types (first 1054 days after installation), 86% of the chart recorders were operating 
for >75% of the interval; 81% of the digital recorders had similar performance 
(Table 2-4). These performance ratings should be considered in management of the 
monitoring network. The initial performance of the chart recorders surpasses that of 
the digital equipment; their longevity is proven; the record is continuous (not point 
observations by time intervals); and, if data retrieval and station maintenance are 
regular, data management procedures can be as automated as that for digital 
recorders. New Steven's chart recorders and platforms should be considered for 
replacement of old instrumentation, especially for remote, seldom-visited sites. 
Most of the existing units are experiencing failure due to decaying installation 
platforms, not necessarily due to failure of the recording equipment. Repairs on the 
chart instruments can generally be made in place without the diagnostic equipment 
needed for digital units. The digital units should be located in the more accessible 
locations, since maintenance frequency is generally higher, recorders are less 
reliable, and diagnoses are more difficult. 
Estimation of Missing Water Level Data 

The swamp hydrology model requires starting water depths throughout the 
swamp and a dataset of bi-weekly, average water depths for model calibration. 



72 



Table 2-4. Summary of water level recorder performance prorated to the initial 
operating period (1054 days). 



Station 


Duration of 

Recorder 

Operation (days) 


Proportion of Days 

Recorder Functioning 

Properly 


Billy's Lake 


1143 


99 


Chesser Prairie 


793 


92 


Coffee Bay 


960 


82 


Cypress Creek 


1027 


89 


Durdin Prairie 


964 


82 


Floyd's Prairie 


1117 


97 


Honey Prairie 


224 


24 


Sapling Prairie 


639 


85 


Sapp Prairie (digital) 


667 


98 


Suwannee Creek (digital) 


1009 


85 


Suwannee River 


1058 


91 


Sweetwater Creek 


878 


76 


Craven's Hammock 


1039 


89 


SCFSP 


715 


68 


Double Lakes 


958 


91 


SCRA 


1017 


97 


Chase Prairie 


1017 


97 


Seagrove Lake 


974 


92 


Kingfisher Landing 


994 


94 


Gannett Lake 


872 


83 


Territory Prairie 


746 


71 


Sill (Brown Trail) 


886 


84 



Table 2-4-continued. 



73 



Station 


Duration of 

Recorder 

Operation (days) 


Proportion of Days 

Recorder Functioning 

Properly 


Moonshine Ridge 


874 


83 


Suwannee Creek (chart) 


854 


97 


Soldier's Camp 


681 


97 


Sapp Prairie (chart) 


629 


60 



Swamp water level recorders and staff gauges provided a partial daily water level 
record, due to malfunctioning recorders; estimates of missing daily data were needed 
to calculate average bi-weekly water depths for the model, and to estimate 
vegetation species-hydroperiod relationships (see Chapter 6). Correlation and 
simple linear regression procedures were used to estimate missing daily data for 
each recorder. During 1980-1995 there were only 6 bi-weekly intervals when all 
recorders were operating concurrently. Therefore comparisons for the best 
correlations and regressions were calculated for reduced intervals only among local 
recorder pairs (the nearest 1-3 stations, using relationships in sequence of highest to 
lowest r^ 2 until a complete daily dataset resulted) for the 1980-1995 missing data. 
All regression pairs met assumptions of linearity, independence and normality of 
residuals, independence of data, and non-autocorrelated residuals (Durbin- Watson 
D) (Myers 1990). Most data pairs were not successive; there were frequent errors in 
the recorded data so that sequential days with recorded data did not occur at every 
recorder simultaneously. Only non-regressed, original recorder data were used in the 



74 
correlation and regression calculations. Several regression relationships were 

necessary for each station to ensure complete data coverage during the interval. 
Missing data for all chart and digital recorder stations were estimated for 1941-1979 
using regression relationships between the stations and the long-term daily staff 
gauge at SCFSP or SCRA (using whichever had the higher correlation) (Myers 
1990). Missing chart station data during 1980-1991 were estimated with regressions 
among chart station data. Since no digital recorders were operating before 1992, 
digital station data for 1980-1991 were estimated with equations developed in 
regressions of chart and digital station data for 1992-1995. Best regression 
relationships between chart and/or digital stations were used to estimate missing 
data at all stations during 1992-1995. Best correlation pairs, regression coefficients 
(P < 0.05) and equations, and missing data estimation intervals are listed in Table 2- 
5. 

Starting water depth and biweekly average water depth were needed to start 
and calibrate the hydrology model. The daily data (actual and regression-estimated) 
at 30 stations were averaged for biweekly intervals during 1941-1993, and 
interpolated using ARCGPJD's (version 7.0, ESRI, Inc., Redlands, CA, 92373) 
circular kriging algorithm to create biweekly water depth estimates in each of the 
model's 10,672 (500X500 m) grid cells. Several of the recorders were at locations 
with topographic, and therefore water depth variability, at a resolution smaller than 
that of the model (500X500 m). Therefore model output was compared to both the 
interpolated data and the original recorder data to determine model performance. In 



75 



Table 2-5. Best correlation pairs and regression equations used to estimate missing 
water level recorder data during 1941-1995. 



Interval and 
Predicted Station 

(Y) 


Predictor Station 
(X) 


r 2 ■ 


Regression Equation to 
Estimate Y 


1941-1979 








Billy's Lake 


SCFSP 


0.9952 


y=1.01883x- 0.655750 


Chase Prairie 


SCRA 


0.8352 


y = 0.474747x+ 19.13962 


Chesser Prairie 


SCRA 


0.9188 


y = 0.919405x + 2.995167 


Coffee Bay 


SCRA 


0.9197 


y=1.177936x- 6.347071 


Cravens 
Hammock 


SCFSP 


0.9685 


y=1.117538x- 3.67679 


Cypress Creek 


SCRA 


0.8646 


y=1.555137x- 22.893495 


Double Lakes 


SCRA 


0.5675 


y = 0.481662x+ 19.95717 


Durdin Prairie 


SCFSP 


0.8856 


y = 0.21442x + 29.80959 


Floyd's Prairie 


SCFSP 


0.9859 


y = 0.636968x+ 13.31684 


Gannett Lake 


SCRA 


0.8518 


y = 0.70466x+ 11.05423 


Honey Prairie 


SCFSP 


0.7635 


y = 0.469929x + 20.665821 


Kingfisher 
Landing 


SCRA 


0.7111 


y = 0.49819x+ 19.33553 


Moonshine Ridge 


SCRA 


0.7427 


y = 0.343284x + 23.25431 


Suwannee River 


SCFSP 


0.7884 


y = 2.331444x- 47.0322 


Sapp Prairie 
(chart) 


SCFSP 


0.7302 


y = 0.306802x + 25.85369 


Sapp Prairie 
(digital) 


SCRA 


0.9536 


y = 0.807310x + 6.93015 


Sapling Prairie 


SCFSP 


0.9869 


y = 0.67846x+ 13.30181 


Suwannee Creek 
(chart) 


SCFSP 


0.4145 


y = 0.509266x+ 18.80288 





Table 2-5-continued 






It 






Interval and 












Predicted Station 


Predictor Station 


r 2 ■ 

* adj 


Regression Equation to 






(Y) 


(X) 




Estimate Y 




Suwannee Creek 


SCFSP 


0.6438 


y=1.037264x + 0.562694 




(digital) 












Seagrove Lake 


SCRA 


0.9190 


y = 0.878125x + 4.538393 






Sill (Brown Trail) 


SCFSP 


0.3881 


y=1.1719092x- 6.25709 






Soldier's Camp 


SCRA 


0.6514 


y = 0.715641x + 8.136278 






Sweetwater Creek 


SCRA 


0.8757 


y = 0.767672x + 6.413319 






Territory Prairie 


SCRA 


0.7243 


y = 0.66398x+ 12.64588 






1980-1991 












SCFSP 


Gannett Lake 


0.7670 


y=1.288821x- 12.483251 








SCRA 


0.7160 


y=1.075264x- 4.308377 








Chase Prairie 


0.6600 


y=1.973633x- 37.028758 






SCRA 


Seagrove Lake 


0.8955 


y=1.006442x- 0.323629 








Gannett Lake 


0.8811 


y=1.452196x- 16.932219 








Chase Prairie 


0.8428 


y=1.73762x- 26.855268 






Chase Prairie 


Territory Prairie 


0.8517 


y = 0.612281x+ 13.896672 








SCRA 


0.8428 


y=1.73762x- 26.855268 








Seagrove Lake 


0.7067 


y = 0.477543x + 18.991341 






Territory Prairie 


Chase Prairie 


0.8517 


y = 0.612281x+ 13.896672 








Seagrove Lake 


0.6235 


y = 0.692281x+ 11.529932 






Double Lakes 


Kingfisher 


0.8672 


y = 0.957434x+ 1.587892 








Landing 


0.6255 


y = 0.381085x+ 24.392432 








Suwannee Creek 












(chart) 


0.6153 


y = 0.654627x+ 13.40319 








Territory Prairie 










Seagrove Lake 


SCRA 


0.8955 


y = 0.890131x + 4.100752 








Chase Prairie 


0.7067 


y= 1. 482 181x- 17.445768 








Territory Prairie 


0.6235 


y = 0.902959x+ 3.326217 







Table 2-5-continued 






11 


p 




Interval and 












Predicted Station 


Predictor Station 


r 2 ■ 

1 idj 


Regression Equation to 






(Y) 


(X) 




Estimate Y 




Sill (Brown Trail) 


Soldier's Camp 


0.4156 


y = 0.888818x + 4.774716 






Suwannee Creek 


0.3987 


y = 0.733971x+ 8.450264 








(chart) 












Sapp Prairie 


0.3642 


y=1.692363x- 26.697464 








(chart) 


0.3596 


y=1.392526x- 13.735103 








SCFSP 










Kingfisher 


Double Lakes 


0.8670 


y = 0.905729x + 3.545421 






Landing 


SCRA 


0.7111 


y = 0.49819x+ 19.33553 








Chase Prairie 


0.6601 


y = 0.953498x + 2.767062 






Gannett Lake 


SCRA 


0.8811 


y = 0.614203x+ 14.379507 








SCFSP 


0.7670 


y = 0.595840x+ 15.974003 








Seagrove Lake 


0.7706 


y = 0.770 187x+ 8.587597 






Sapp Prairie 


Soldier's Camp 


0.7622 


y - 0.377933x + 23.607006 






(chart) 


Gannett Lake 


0.5655 


y = 0.526923x+ 17.55543 








SCFSP 


0.5616 


y = 0.26444x + 27.336647 






Soldier's Camp 


Sapp Prairie 


0.7622 


y = 0.377933x + 23.607006 








(chart) 


0.6343 


y = 0.61218x + 0.68960 








Gannett Lake 


0.6432 


y = 0.802501x + 5.000845 








Suwannee Creek 












(chart) 


0.5300 


y = 0.636551x+ 11.030303 








SCRA 










Moonshine Ridge 


Gannett Lake 


0.5700 


y = 0.46871x+ 19.728948 








SCRA 


0.4873 


y = 0.353624x + 22.878873 








Seagrove Lake 


0.3803 


y = 0.30953x + 24.463134 






Suwannee Creek 


Soldier's Camp 


0.6432 


y = 0.81 1783x + 8.608739 






(chart) 


Double Lakes 


0.6309 


y=1.702882x- 27.323 189 








Kingfisher 


0.5284 


y=1.317963x- 12.913351 








Landing 










Coffee Bay 


Seagrove Lake 


0.9200 


y=1.16484x- 5.94484 








SCRA 


0.9197 


y=1.177936x- 6.347071 








Chase Prairie 


0.7600 


y = 2.263836x - 45.920005 





Table 2-5--continued. 



78 



Interval and 








Predicted Station 


Predictor Station 


-2 » 


Regression Equation to 


(Y) 


(X) 




Estimate Y 


Chesser Prairie 


Seagrove Lake 


0.9863 


y=1.044685x- 1.672174 




SCRA 


0.9841 


y = 0.998382x + 0.09812 




SCFSP 


0.9648 


y = 0.819478x + 7.957763 


Durdin Prairie 


SCFSP 


0.8881 


y - 0.207952x + 30.030537 




Territory Prairie 


0.7600 


y = 0.43822x + 21.128403 




Double Lakes 


0.5960 


y = 0.5301 82x+ 16.902973 


Sapling Prairie 


SCFSP 


0.9233 


y = 0.679905x+ 13.257842 




SCRA 


0.9083 


y = 0.877289x + 4.960941 




Double Lakes 


0.6800 


y=1.77944x- 3 1.546924 


Billy's Lake 


SCFSP 


0.9952 


y=1.018813x- 0.655750 




SCRA 


0.8815 


y=1.178149x- 8.135219 


Suwannee River 


SCFSP 


0.7884 


y = 2.331444x- 47.0322 


Craven's 


SCFSP 


0.8662 


y=1.095096x- 2.897342 


Hammock 


SCRA 


0.7525 


y=1.254816x- 10.514335 


Floyd's Prairie 


SCFSP 


0.9280 


y = 0.629658x+ 13.569917 




Seagrove Lake 


0.6400 


y = 0.845072x + 4.604841 




Territory Prairie 


0.6300 


y = 0.935060x+ 1.052165 


Suwannee Creek 


SCFSP 


0.6438 


y=1.037264x + 0.562694 


(digital) 


Double Lakes 


0.4471 


y = 2.74672x- 68.9186 


Honey Prairie 


SCFSP 


0.7205 


y - 0.42043x + 22.383008 




Seagrove Lake 


0.6200 


y = 0.520723x+ 17.994814 


Sweetwater Creek 


Seagrove Lake 


0.6100 


y = 0.870450x + 2.578536 




Gannett Lake 


0.5100 


y = 0.657803x+ 10.309298 




SCRA 


0.4878 


y = 0.61605x+ 11.963422 




SCFSP 


0.4703 


y = 0.50043x+ 17.004948 


Cypress Creek 


SCRA 


0.8646 


y=1.555137x- 22.893495 




SCFSP 


0.8501 


y=1.292408x- 11.177002 




Seagrove Lake 


0.8300 


y=1.672840x- 27.301 101 


Sapp Prairie 


SCRA 


0.9536 


y = 0.807310x + 6.93015 


(digital) 


Seagrove Lake 


0.9200 


y = 0.78096x + 7.837301 




Chase Prairie 


0.8300 


y=1.627919x- 23.007385 



Table 2-5-continued. 



79 



Interval and 








Predicted Station 


Predictor Station 


r 2 » 

• adj 


Regression Equation to 


(Y) 


(X) 




Estimate Y 


1992-1995 








Sill (Brown Trail) 


Suwannee River 


0.9523 


y=1.196004x- 6.83 1348 




Billy's Lake 


0.7369 


y = 2.369154x- 48.318059 


SCFSP 


Billy's Lake 


0.9955 


y = 0.979881x + 0.701561 




Floyd's Prairie 


0.9245 


y=1.479422x- 17.678245 




Suwannee River 


0.8158 


y = 0.38274 lx + 21.745865 


SCRA 


Chesser Prairie 


0.9895 


y = 0.976399x + 0.824623 




Coffee Bay 


0.9197 


y = 0.782395x + 7.836545 


Chase Prairie 


Sapp Prairie 


0.8266 


y = 0.510426x+ 17.923383 




(digital) 


0.7582 


y = 0.336781x + 24.149895 




Coffee Bay 






Territory Prairie 


Durdin Prairie 


0.7592 


y=1.74278x- 28.079932 




Floyd's Prairie 


0.6279 


y = 0.678246x+ 12.817394 




Billy's Lake 


0.6024 


y = 0.449532x + 21.23899 


Double Lakes 


Sapling Prairie 


0.6819 


y = 0.387576x + 24.193191 




Durdin Prairie 


0.5957 


y=1.137446x- 3.929222 




Floyd's Prairie 


0.5289 


y = 0.419301x + 23.597785 


Seagrove Lake 


Chesser Prairie 


0.9863 


y = 0.9444 17x + 2.070188 




Sapp Prairie 


0.9207 


y=1.178908x- 6.328907 




(digital) 


0.9115 


y = 0.0784023x + 7.845797 




Coffee Bay 






Kingfisher 


Double Lakes 


0.8670 


y = 0.905729x + 3.545421 


Landing 


SCRA 


0.7111 


y = 0.49819x+ 19.33553 




Chase Prairie 


0.6601 


y - 0.953498x + 2.767062 


Gannett Lake 


Coffee Bay 


0.6794 


y = 0.60133x+ 14.705007 




Chesser Prairie 


0.6401 


y = 0.763805x + 8.822456 




Sapp Prairie 


0.6218 


y = 0.958325x+ 1.849815 




(digital) 






Sapp Prairie 


Gannett Lake 


0.5655 


y = 0.516923x+ 17.55543 


(chart) 


SCFSP 


0.5616 


y = 0.26444x + 27.336647 




Seagrove Lake 


0.4273 


y = 0.360912x + 23.374758 



Table 2-5-continued. 



80 



Interval and 








Predicted Station 


Predictor Station 


r 2 * 

1 «dj 


Regression Equation to 


(Y) 


(X) 




Estimate Y 


Soldier's Camp 


Gannett Lake 


0.6343 


y = 0.91218x + 0.68960 




SCRA 


0.5300 


y = 0.636551x+ 11.030303 


Moonshine 


Chesser Prairie 


0.7489 


y = 0.477916x+ 18.287191 


Landing 


Sapp Prairie 


0.6942 


y = 0.561003x+ 15.322119 




(digital) 


0.4970 


y = 0.364872x + 22.377580 




Coffee Bay 






Suwannee Creek 


Double Lakes 


0.6256 


y=1.659745x- 27. 11490 


(chart) 








Coffee Bay 


Chesser Prairie 


0.9308 


y=1.276304x- 9.981905 




SCRA 


0.9197 


y=1.177936x- 6.347071 




Sapp Prairie 


0.9080 


y= 1. 5 193x- 18.647531 




(digital) 






Chesser Prairie 


Seagrove Lake 


0.9863 


y=1.044685x- 1.672174 




SCRA 


0.9841 


y = 0.998382x + 0.09812 




SCFSP 


0.9648 


y = 0.819478x + 7.957763 


Durdin Prairie 


SCFSP 


0.8881 


y = 0.207952x + 30.030537 




Sapling Prairie 


0.7719 


y = 0.344244x + 24.585459 




Territory Prairie 


0.7600 


y = 0.43822x + 21.128403 


Sapling Prairie 


SCFSP 


0.9233 


y = 0.6799 15x+ 13.257842 




SCRA 


0.9083 


y = 0.877289x + 4.960941 




Craven's 


0.8192 


y = 0.437932x + 21.542202 




Hammock 






Billy's Lake 


SCFSP 


0.9952 


y=1.018813x- 0.655750 




Floyd's Prairie 


0.9215 


y= 1. 40 1629x- 14.917255 




SCRA 


0.8815 


y=1.178149x- 8.135219 


Suwannee River 


SCFSP 


0.7884 


y = 2.331444x- 47.0322 


Craven's 


SCFSP 


0.8662 


y=1.095086x- 2.897342 


Hammock 


Billy's Lake 


0.8372 


y=1.226685x- 7.493928 




Sapling Prairie 


0.8192 


y=1.880079x- 34.234817 


Floyd's Prairie 


SCFSP 


0.9280 


y = 0.629658x+ 13.569917 




Billy's Lake 


0.9226 


y = 0.658218x+ 12.578014 




Seagrove Lake 


0.6400 


y = 0.845072x + 4.604841 





Table 2-5~continued. 


81 






Interval and 

Predicted Station 

(Y) 


Predictor Station 
(X) 


r 2 ■ 

r adj 


Regression Equation to 
Estimate Y 




Suwannee Creek 
(digital) 


Craven's 
Hammock 
Sapling Prairie 
SCFSP 


0.8432 
0.6972 
0.6438 


y = 1. 09583 lx- 1.956902 
y = 2.142633x- 42.481572 
y=1.037264x + 0.562694 




Honey Prairie 


Sapp Prairie 
(digital) 
Chesser Prairie 
SCFSP 
SCRA 


0.9326 
0.9248 
0.7205 
0.6566 


y = 0.926357x + 3.313969 
y = 0.688285x+ 11.898168 
y = 0.420434x + 22.383008 
y = 0.499013x+ 18.817358 






Sweetwater Creek 


Seagrove Lake 
Chesser Prairie 
Sapp Prairie 
(digital) 


0.6100 
0.5997 
0.5685 


y = 0.87045x + 2.578536 
y = 0.819802x + 4.504097 
y = 0.925475x + 0.768341 






Cypress Creek 


Sapp Prairie 
(digital) 
Chesser Prairie 
SCRA 
Seagrove Lake 


0.9467 
0.8669 
0.8646 
0.8300 


y - 2.222674x - 46.997072 
y=1.815132x- 32.417015 
y=1.555137x- 22.893495 
y=1.67284x- 27.301 101 






Sapp Prairie 
(digital) 


SCRA 

Cypress Creek 
Chesser Prairie 


0.9536 
0.9467 
0.9226 


y = 0.80731x + 6.93015 
y = 0.426513x + 21.941910 
y = 0.789828x + 7.541488 




1 


' All regression relationships were significant at P < 


[).05 or less. 




most cases model performance that corresponded better with the original recorder 


data represented stations with detail at a resolution smaller than the 500X500 m cell 


size of the model (e.g., the recorder was located in a 


ditch, small lake, or stream 


bed). Recorder data estimated with regression equati 


ons and interpolated for model 


performance evaluations are plotted in Figure 2-2. 





82 
Correlative relationships that permit missing data estimation also indicate 
redundancies in the water level recorder network. Although these redundancies 
were useful in model database development and estimating missing data using 
regression relationships, they represent a significant investment of personnel 
required to maintain the recorders and database. There are several approaches to 
eliminating redundant stations in the recorder network. Selecting stations depends 
on the intended use of the recorder data. If the interest is in representing local 
uniqueness while eliminating redundant stations, 17 stations should be maintained. 
These include unique stations (Figure 2-1; Craven's Hammock, Sweetwater Creek, 
Durdin Prairie, Double Lakes, Kingfisher Landing, Moonshine Landing, Sapling 
Prairie, Floyd's Prairie, Suwannee River, Suwannee Creek, Gannett Lake, Cypress 
Creek, Soldiers Camp, Territory Prairie, Chase Prairie) and redundant stations which 
could be used to estimate missing water elevations at other stations (SCFSP or 
Billy's Lake, and Seagrove Lake or Chesser Prairie). If the interest is in maintaining 
the recorder network for the best predictions of missing data, there are 17 stations 
that should be maintained. Stations highly correlated (r^ 2 > 0.90) with at least one 
other station should be maintained, as well as those most unique (not highly 
correlated with at least one other station). Stations having highly correlative 
relationships with other stations include SCRA, Seagrove Lake, Sapp Prairie 
(digital), Suwannee River, Floyd's Prairie, and SCFSP. Those that are most unique 
( r adj 2 < 0-90) include Chase Prairie, Craven's Hammock, Durdin Prairie, Double 
Lakes, Gannett Lake, Kingfisher Landing, Moonshine Ridge, Soldier's Camp, 



83 
Suwannee Creek (digital), Sweetwater Creek, and Territory Prairie. If the interest is 
in eliminating redundancy to minimize resources needed to represent temporal and 
spatial variability in water surface elevations, 13 stations should be maintained. 
These are SCRA (r^ 2 > 0.90 with Chesser Prairie, Coffee Bay, Sapling Prairie, 
Seagrove Lake), SCFSP (r^ 2 > 0.90 with Billy's Lake, Floyd's Prairie, Chesser 
Prairie, Craven's Hammock, Sapling Prairie), Sapp Prairie digital (r^ 2 > 0.90 with 
Cypress Creek, Chase Prairie), Suwannee River (r^ 2 > 0.90 with Brown Trail-Sill), 
and the unique stations at Durdin Prairie, Double Lakes, Territory Prairie, Gannett 
Lake, Kingfisher Landing, Moonshine Ridge, Soldiers Camp, Suwannee Creek 
(digital), and Sweetwater Creek. Time and resources saved by eliminating network 
redundancies should be invested in improving recorder performance at the remaining 
stations. Without these redundant stations, missing data estimations with the 
recorder relationships in Table 2-5 will not be possible. However, biweekly water 
level fluctuation estimates can be made with the swamp hydrology model (see 
Chapter 3), and model performance accuracy can be assessed with data recorded at 
the remaining stations. 

Precipitation Gauge Network Assessment 

Background 

Considerable refuge resources are devoted to maintenance of precipitation 
recorders and management of retrieved data. These data are used to estimate area 
daily rainfall. It was uncertain how representative these recorder data were of actual 



84 
daily area rainfall, because the variability of daily rainfall had not been examined. It 
was possible that variability in daily area rainfall exceeded network resolution for 
daily area rainfall calculations, and that these data would be more appropriately 
summarized over longer periods (weeks to months) to estimate average daily rainfall 
throughout the swamp. 

Accuracy of the precipitation recorder network in estimating area rainfall 
was assessed using a technique that compares variation in precipitation measured at 
recording stations, weighted by the area of coverage for the measurement, and 
adjusted by the spatial and relative variances and covariances calculated among 
stations and within the watershed (Still and Shih 1990, Shih 1982). Total 
precipitation variation in a watershed is due to variation at individual stations and 
variation among stations, or spatial variability, as 

S 2 (x) = S r 2 (x)+S c 2 (x) 

where 

s 2 (x) = total variance of mean rainfall. 
s r 2 (x) = relative variance of mean rainfall, and 
s c 2 (x) = spatial variation of mean rainfall. 

Relative variance is dependent on the network density; it can be reduced by 

increasing the density of recording stations (Shih 1982). Relative variance can be 

calculated to represent randomly placed stations (Method A), or stations allocated 

randomly among strata (Method B), or located relative to within-stratum variability 

(Method C): 



85 



Method A s r 2 (x) = (l/N)(s 2 -s okl ) 

Method B s r 2 (x) - 1 (Wi'/N,) (s 2 - s okll ) 
i=l 

n 

Method C s r 2 (x) - (1/N) [ I (WO {(§„,- s okh ) 2 } m ] 

i=l 

where 

N = the total number of stations 

Nj = the number of stations in the rth stratum 

s 2 ■ the average variance within the watershed 

s okl = the average covariance within the watershed 

i, j = the stratum 

n = the number of strata 

Wj ■ the area ratio for the /th stratum 

s^ 2 = the average variance within the rth stratum 

Sokii = me average covariance within the rth stratum. 

Spatial variance reflects a characteristic of the watershed and does not necessarily 

decrease with increasing network density (Shih 1982). It is affected by within- 

stratum covariance and the proportional area of the stratum relative to the watershed. 

Station placement should maximize stratum homogeneity; spatial variance should 

decrease if each stratum represents homogeneous areas of rainfall. Increased 

precision in rainfall estimates should then result with non-random placement of 

gauges in the watershed (Methods B and C) (Still and Shih 1990, Shih 1982). 

Spatial variance is calculated by 



Method A s c (x) = s, 



okl 



Methods B and C s c 2 (x)= I (W 2 )(s okli ) + 2 £ £ {(W, WjXW) 

i=l i=l j=l 



86 
where: 

s kiij = the average covariance between the rth andyth strata. 
These parameters describe the variability of watershed precipitation. Based on the 
total variance of the mean rainfall, the number of stations required to estimate mean 
rainfall within a desired statistical accuracy can be calculated (Still and Shih 1990, 
Shih 1982). For random gauge placement (Method A) and specified a, P, and mean 
(x) rainfall within the watershed, 

N = t a 2 {(s 2 -s okl )/(px) 2 }. 
For stratified gauge placement (Methods B and C), 

N = [{t a lW i (s oi 2 -s okli ) 1/2 }/{PLW i x I }] 2 , 
H i=l 

where x ; is the mean rainfall within the rth stratum. Accuracy of the gauging 

network is proportional to the variability of rainfall at each station, the size of the 

basin represented by the gauge, and the amount of variability among stations (Shih 

1982). The proportion of gauges to be allocated to each stratum can be calculated 

based on a weighted ratio, 

N, = NQ, 

where 

C, = {W, (s^-s^f 2 } / {£ Wi (s^)}. 

i=l 
Stations can be relocated to better represent the spatial variance of rainfall in the 
watershed, based on this ratio. 



87 
Methods 

Two subsets of precipitation recorder stations were selected for assessment 
of network accuracy. One subset (subset 1) contained all stations in the network 
with daily precipitation data for at least half of the recording days during 31 March 
1992-3 July 1995. This subset represented a recorder network with 14 stations 
(Chase, Territory, Floyds, Durdin, Sapp, and Sapling Prairies, Kingfisher Landing, 
Double Lakes, Seagrove Lake, SCRA, SCFSP, Cravens Hammock, Coffee Bay, and 
Suwannee River). The second subset (subset 2) eliminated 3 stations from subset 1, 
to contain only those stations that were easily accessed; this subset represented a 
network with 1 1 stations (Chase, Territory, Durdin, and Sapp Prairies, Kingfisher 
Landing, Double Lakes, Seagrove Lake, SCRA, SCFSP, Coffee Bay, and Suwannee 
River). Daily precipitation data were averaged for each station by daily, bi-weekly, 
and monthly intervals (Table 2-6), i.e., daily average precipitation was calculated by 
summarizing over days, bi-weekly, and monthly periods. Interpolated surfaces of 
daily averages calculated from daily, bi-weekly, and monthly average data were 
created using the ARCVTEW inverse-distance-weighted (IDW) procedure, and 
contoured at 1 mm. Several contouring intervals were calculated for each subset and 
average to determine minimum differences among interpolated stations; intervals 
less than 1 mm resulted in partitioning the stations into single station strata (i.e., 
each station was isolated), whereas intervals greater than 3 mm resulted in no 
contours, i.e., all stations belonged to 1 stratum. This means that measured 
differences among stations in average daily rainfall volume calculated by daily, 



• *-* 



-o os 

£ — 

co >> 

13 

s w 

■ 

1 £ 
co On 

<U — 

£ xi 

s a 

co 

oo 

c 



"O 

1) 

3 

E 



to 5 

C c« 

O <U 

•5 60 

$ s 

a > 

'o * 

<l> >. 

2 ° 

1) CO 

Z c 

CO O 

£M 

so ° 

1| 

H E 



t = J 



|*«I 



14 I 

2 = 5 




o - e 

fill 




lit 



00 
00 



os 



r- 


■* 


ID 


o 


o 


o 


o 


■*■ 


SO 


00 


00 


r^ 


r^ 


r~ 


SO 


SO 


</"> 


r- 



o 

SO 

ON 



§ 2 £ 2 
2 IZ « no 



O 


t^- 


r- 


00 


o 


r- 


(S 


5 


^ 


o 


Os 


vO 


f) 


— 1 




o 


o 


00 


00 


00 


r- 


r- 


*0 


0s 



o> 


P4 


o 


00 


on 


0\ 


<s 


m 


■f 


<N 


<N 


<N 









Os 


o 


O 


o 


O 


o 


o 



<s 




00 

CI 


00 


00 




r- 

m 


Os 




o 


O 


O 


o 


o 


© 


O 


o 


o 



If 

H 1 



J 



© vO r*> © ON 
rn <N r<l r-» fS 

©'©'©' O © 



88 



Os 



1*1 



Os 


00 


o 


SO 


r- 


00 


_* 


00 


TT 


o 


CI 


fS 


C*l 


m 


fS 


CS 


It 


<s 


m 


tN 


o 


© 


© 


O 


o 


o 


© 


o 


o 


O 



II 
















ll 


•t 


so 


SO 


_ 


O 


o 


rr 




CI 


00 


<S 

© 


00 


00 


o 


00 
1/1 
CI 



p rr so 00 Os tN <N 

ob ^ on r*" ci so* fW 

O Tj- «rt ON Tf CN t«- 

<S tS fS — fS — c fS 









PQ 


-a 
£ 


cd 
E 

0. 




.s 


(A 


o 


i3 


o 


U 


Q 


u- 



a 

: i » 

g - c 

o- S 

» *f 

P. E ap 



? 

2 



t/3 



I 



U 3C 









5 



J 



I 
t 



CO 

•n 
J 

<u 

g 

-c 





i> 




«3 


■i 




a. 


J 




^ 


SJ 




o 


o 


•j- 






r^l 


H 


8P 


u. 


cu 


a> 


U 


H 


c« 


cyj 



89 
biweekly, or monthly periods are generally less than 1-3 mm. These interpolated 

surfaces provided groupings of stations within strata; for each interval, stations with daily 

averages that were within 1 mm of each other were grouped in the same stratum. 

Groupings of stations among strata, and strata area, variances, covariances, required 

number of stations for specified network accuracy, and allocation of gauges among strata 

are listed in Tables 2-7 and 2-8. Strata delineations for daily, biweekly, and monthly 

networks of 1 1 and 14 stations are given in Figures 2-3 through 2-8. 

Results of Precipitation Network Analysis 

Relative variance of the recording network is dependent on the number of 

stations, and how recorders are partitioned among strata. In the 14 gauge network the 

relative variances of daily averages from daily data calculated by Methods A, B, and C 

were 0.0364, 0.0463, and 0.0437, respectively. The similarity in these variances suggest 

that either method (random or stratified) would be appropriate for relative variance 

calculation, since within-stratum variability is low. Therefore, stratified allocation of 

gauges affords no increase in precision of relative variance over randomly sampling the 

watershed. This was also true of daily averages calculated with bi-weekly and monthly 

data (Table 2-8). However, there are differences in spatial variance of daily data 

calculated by methods A, B, and C. The random sample method (A) results in slightly 

higher spatial variance for daily rainfall estimates from daily data (0.4937) than the 

stratified sample (0.3812). This suggests that rainfall distribution is not homogeneous 

throughout the watershed, and that the network benefits from stratification. This 

difference does not occur when the daily data are summarized over bi-weekly and 



90 



es 3 



■£ v as 
55 






cs 

> 
o 

u 

W) 
CB 

s. 



R 

u 

— 

e 



00 

o 



1^- 

o 



o 

<N 



O 



(N 



O 



© 



<N 






1 
<W3 



<tfl 



CO 



1 
1 


© 


0.47 
0.3 


00 






o 




o 



r<i 



O 



©' 



— « r- 

o — i 



oo 



m o 



N O »^0 V> 00 

m i—i m «— i r<"> t*> 



O O ~* ** t-i *■* 



o — 



oo on t 
O >0 i— vO 

1-- O — ' O 






© 
©' 



«tfl 



B 
o 

•a 

-— 



1 1 s 
111 

■C tfl ** 
Z u 



2 -a 

II 

W Q 



f-sl-8 

J -a £ J 

Jgf «& 

3 a .S -o oa 

age S?^ 

M Q H E on co O 



<D 


44 


.#• 


s 


o- 


E 


(U 


B 


c 


I 


c 


c/l 


«1 


H 


£ 


B 





u 


« 


.3 


'C 


a 


at 




t_ 


<u 


Oh 


.> 


<U 


1 


t/V 


9 


JB 


a 

i> 


u 


CO 



y 

CO 



:s I 4 

So" 

Oi « «i 
a-- $22 

^s6 



4L 3 



(N 



r^ 



*S5 



91 



(7> 






00 



SO 

6 



o 



u 

e 

.2 

S 

u 

v 

DC 

08 

U 

o 
- 



S3 
S- 

B 
V 

PQ 



SO 

O 



o 



m 



CM 

O 



rn 

o 

o 



CM 



CO 

O 



5f 
o 



o 
SO o 

o 
© 



o 



</-l 



CM 

o 






CM 

O 



ON 

o 



s 



1 
<«5 



o 
o 



o 



o 



5 
o 



SO 00 

o o 
o o 



so os so so in 

•■* ^^ *-M ^-« »— H 



o — 



00 OS O 00 t^ 
O O CM O O 



o 



ooooo ooooooo 



S 

e 

<W5 



SO 

© 



O 



oo 
o 

o 



3 
3 

c 
o 

O 



CM 

H 



e 
- 




•c 

I 

a. 



> 

2 



B 

^ D h fa u O c/3 co U co tvo oo 



Dm 

u 



CQ 






3 o (J u ea wq 



•2 

'C 

J 

00 



a. 
m 

oo 



o * a 

ijj 



CM 



cn 



I! c 
i I 



92 



3 

a 

c 
o 



CN 



£ 


"9- 






C-l 

O 


•/-> 
■* 






o 






- 


© 






O 


d 






©' 






d 


£ 
























tratu 
Area 
(ha) 








§ 


I 






ON 






90 
<N 

00 


OB 


f~ 






c> 


r~ 






*~* 






"™* 


6 

e 
•S 

■ 
























cn 
O 

© 






s 

O 


i 






s 

O 








> 
























o 
























U 
























■j 

■ 


O 






i 


s 






© 


<N 




ON 


E 


o 






© 






© 






o 


► 
























•< 
























■ 
























** 
























08 
























E 
























■** 
























c/) 
























a 


1 
1 






s 


o 






o 








i 


1 






o 


o 






o 








v 
























V 
























BQ 


ss 






So 


m 

O 


5 
o 




o 




m 






© <=> 






© ° 


o 


© 




1 

o 


— 




1 © 


._ 


3 






5 


en 












ON 


3 




© 






o 


o 












CO 


«/3 


o 






o 


o 






o 






O 


w_ 


o 






§ 


«n 






On 






r^ 


1 
OB 


*— ' 






o 






«— * 






CN 


o 






o 


© 






© 






— ' 




v© O </") 


00 


-» r» 


vO NO 


*Ti 


r- «* 


</-> 


ON 






CN O On 


<S0 


o — o 


o 


*-H f— -i 


o o 


© 


o o 


o 


— H 






r«-> — < co 


©' © © 


o 


© © 


© o 


o 


o © 


© 


© 






-" — ' ~* 


3 






















t> 


■ 

<«5 






















d 


e 






















g 


m 






















^H 


■ 
e 
•c 
■ 


to .5 .2 

J tj £ 


(LI 

s 

Oh 


.s 5 

2 1 


6 
> 

2 

u 

<u 
c - 


<L> 

T 
1 

0h 


*g 

«J .a 

CQ 3 -~ 


1 

<L> 

> 








u 

00 ■ — 

a <l> .5 
|;c g 

u Dh >, 


£ 


o 


C/3 t/j 


C Oh 




(U O. "5 


o 


2 

U 

O0 






J c o 




2 £P b t 


^3 fl 


CO oo 

3 U 
(73 C/3 


UUtn3 


feb 

o 

00 






«a T3 "c 
§> a c 

M Q H 


jf «r 
























k i ■? 


















"«T 


& 




Estim 
Stra 


CN 






m 


<«■ 






yr\ 







93 



T3 

D 
3 
fl 

'•♦3 

C 

o 






£ 


0.01 
0.87 

0.93 


Stratum 
Area 
(ha) 


2333 
139818 

149032 




e 

.2 

■ 
> 
o 

U 

B) 

I 
E 

V 

> 

ett 

■ 
E 

B 
V 
V 

* 

■ 


0.49 
0.45 

0.37 0.45 

1 2 

0.07 


a 

J 


en o 

© d d 




1.35 
0.84 

0.11 




«/■> o oo oo r~- m o\ -^- vo oc © oo o\ o\ r>- — < oo 
co o © r>- r^ — < m so •— ■ © r» © © o © — © 

— i d «—' © © <-< © © ©©©'©©'©© © © 


3 

e 

<C/3 


d 


e 
<«5 


©' 


■ 

e 
■ 

mm 

m 


SCFSP 

Double Lakes 
Chase Prairie 
Suwannee River 
Coffee Bay 
SCRA 

Seagrove Lake 
Sapp Prairie 
(digital) 

Kingfisher Landing 
Double Lakes 
SCRA 

Seagrove Lake 
Chase Prairie 
Coffee Bay 
Sapp Prairie 
(digital) 
SCFSP 
Suwannee River 


Network, 

Estimate, 

Strata 


2 
3 

N=ll, 

Biweekly 

1 



0> 
3 
C 

C 

o 
o 



1 



t* 


o 




O 






© 


© 




©• 




o 


© 


lit 


fi 




<N 








l|e 


o 




IB 




S 










W^ 




Tl" 


c> 


% 


tm 




™" 




~" 


~* 


V 





























i 














i 




























E 














■ 














> 














o 














u 














tl 






o 




i 


3 

O 


I 

la 


1 
1 


<N 


© 




© 


V 


l 












► 














< 














■ 














■*» 














OS 














E 














■** 














09 














a 














V 














8 














^ 














E 












i 


CO 


fc. 




o 




m t|- 


i 




o 






o o 


© 




o 


— m 


! ° 




O O 


© 


H 


^ — 




en 




m 




..0 


— -" 




o 




o 




•SB 


o 




o 




© 


o 


M — 


m 




r» 




<o 


ON 


*o 


•— 




o 




o 




«x> 


© 




o 




© 


© 




On VO 




v© © m 


oo 


<n m r^ ^ so o 


On 




O — 




o — o 


o 


© © © © o o 




© o 




© o o 


© 


o © © o ©' © 


© 


3 






m 








e 
<«5 






© 














r» 








o 






o 








<W 






o 








■ 

e 
•a 

I 


4> c 

SI 




ao 
M .5 .2 

1— 1 u. Oh 


;£ 

'S3 


4) .9 -t 

<s _ ea .3 o 




aS 


c o 




J2.ce 


o 


„, O D C c3 — C 


2 

U 
on 








•fill 

3 M H 

o c 3 

Q2D 


'E 

<u 
H 




L ■*■* flB 














Estim 
Stra 


(N 


i I 






(N 


ro 



95 



c 




o 




■a 




3 








a. 






f/1 


o 


« 


fi 


OJ) 


a 


3 


o 


so 



•S a 

a .2 

•»- *-» 

-b « 



H I 

e a. 

§ <~ 
•n ° 

«J B 

> O 

u 

■g « 

co eg 

_-T "3 

3 fi 

■Jjj t/1 

S -o 

* C/l 

u u 

> 60 

3 1 

*^ 
03 

i **3 

<N .2 

<U .B 

H -a 



*.tfU 
















nthl 
mat 
hod 


o 
© 


© 


© 




S M N ■» 


<N » O O 

« N - - 




o c *» 
« w 5 


© 


© 


© 






© © © © 




►»*■ 
















nthl 
mat 
hod 


in 

© 
© 


© 


© 




5 M N f 


(S T O vO 
m <N — — 




3 ~ S 
* W S 


© 


© 


© 






© © © © 




>> «T ^ 
















MonthI 
Estimat 
Method 


© 
© 


o 


© 




SS •■* 


r*i »r» © vo 

n N N « 




© 


© 


© 






© © © © 




^u'U 
















iweekl 
stimat 
ethod 


© 
© 
©' 


© 
© 


© 

© 




^ o •* - 

^ a \o t 


m >o ™ b 

m <N (S — 
©'©©'©' 




"MS 
















!►> C ffl 
















iweekl 
stimat 
ethod 


© 
© 

©' 


© 
©' 


© 

© 




J_ O >£> rf 


ui vo —■ C- 
n N N - 

© ©' ©' ©' 




— '*. 7- 
















b« 
















eekl 
mat 
hod 


© 
© 


© 


© 




22 m «- >n 


wi r^ — t- 
m <s ts — 




•5 1 I 


© 


© 


© 






© © © © 




BUS 
















tfW 
















aily 
mal 
hod 


© 


00 


rr 




ffl °"> — I-. 

00 5 — m 


O <N — f* 

Tf 1ft O - 




Q 3 J 


© 


© 


© 




^ ~ °° «"> 


— ©' ©' O 




fl S 
















rf ■ 
















aily 
mal 
hod 


•ft 

O 


00 


n 

** 




n c* _ k. 

00 3 — cs 


O. JS —> r»^ 

^t in vo — 




Q 1 J 


© 


© 


© 




n" 00 " 


—'©'©' O 




w 7- 
















C< 
















Daily 
stimal 
ethod 


s 

©' 


■* 
©' 


Ift 

©' 




1735 
1008 
644 
419 


M oo oo r- 
— rr F ■q- 

— i oo \o m 

—'©'©'© 




- S 
















B 
i 
■ 


■ 

2 


2 






II II II II 

z z z z 


11 II II II 

ca. co. co. co. 


fa> 


IS u 

r 

■ 
z 


o 


* 


t 


B 


■ft © *ft © 


m _» «k a 


e 


a 






e8 


© — 1 — «N 


"ft © •/"> © 


a 


N=14 

Relative Varianc 
Rainfall 


■ 

g 

1 

> _ 

•a 3 
1.1 


O 

4> 

o 

s 

11 


s 

E 

u 

< 

E 
e 


©'©'©'©' 
¥ ii T ii 
a a a « 

©* ©" ©* ©" 

©'©'©'© 
ii ii n T 
c2. co. ca.es. 


© — — cs 
©'©'©'©' 
II II II II 
a a a a 

•«t V T* "♦" 

7 T T T 
z z z z 


Strata Allocatio 
Gauges: 




O. 3 

c/j as 


H M 


Z 


— i- ^- u. 

<2 <2 <2 <2 


t_ l_ t_ t- 



96 







fcrfU 






























ail 








rs 














^ — 


m co 


in 




I E J 








§ 


o 




o 






;v © on >/■> 
g n© o-> »S 


o 


M — 


— 




111 

2 M S 








© 


o 




o 








© 


o © 


© 




>. tf « 






























MonthI 
Estimat 
Method 








i 

d 


cn. 

O 

d 




o 
© 






g © On <r> 

X >0 n N 


d 


fi 00 
fN — 

d d 


>n 

d 




;►» * < 






























£ 1 "2. 


_. 


r~ — <n 


H 


1 






•3- 








I* 


CO CN 


CO 




1 E 1 


ii 


II II II 


3 


o 




o 






ss o *> ■+ 


«*>. 


<N fS 






O 9 « 

3 I J 

2 H S 


z z z z 


O 


d 


o 




© 








© 


O © 


© 




.£> tfU 






























3 « - 
| Ej 








5 


o 




v© 
© 






£? ~ — " n© 

* "" t»> Tf 


^9- 




© 




I! i 








© 


© 




O 








O 


O O 


© 




■ w S 






























>> g> SB 






























eekl 
mat 
hod 








i 


§ 




s 






JO *~ — n© 
<* — r~ tt 






© 

CNl 




£ I 1 








d 


© 




© 








o 


o o 


© 




■ w S 






























*%«?*< 






























Biweekl 
Estimat 
Method 


II 


II II 




1 

d 


NO 
O 




o 






S "" t* «-> 




<^ fS 


© 




z z z 




© 




o 








o 


© © 


© 
































«'U 






























hit 

•5 E 2 








£ 


__ 




n© 






2S 2 "« <n 


o 


m © 


<N 










o 


rn 




cn 






3 £ °° •> 


Tt 


— On 


t^ 




o '5 I 








o 


© 




© 






"• 


— © 


© 




w I 






























f ■ 






























>> 1 "2 

re E j: 








$ 


CI 










» N - -j 


On 


m © 

— On 


fS 




a| 1 








o 


© 




© 






— < 


— © 


© 




w S 






























tf< 






























Daily 
stimal 
ethod 


7 

z 


f s 

Z Z 




8 

d 


© 




© 






S 3 B £ 
3 2 * K 


On 
On 


W1 Tf 

m © 

— on 

— ©' 


© 




« 7 




























■d 




«T 








§ 


a 

I 

"8 

1) 

Q 

§ 

a 
> 

1 

o. 

60 


















9> 

5 

Q 

i 

i 

oo 




E 
1 








ative Variance of M< 
nfall 




a 






ii ii n ii 

z z z z 


II II II II 

CO. CO. CO. CO. 




Network, Vari 
Accuracy 


CO 


<N m T 

ra rt rt 


in 
1 


1 


s 

o 

B 

a 

a 
> 

1 

o 
H 


8 
-*- 

B 


w 

£ 

s 

B 
w 

< 

E 
e 

E 


m" o" wT ©" 
© — — r-i 

? ? 1 1 

b a a a 

o" ©* o" ©" 

©odd 
II II II II 

CO. CO. CO. CO. 


«/-T o" <W ©" 
© — — <s 
©' d ©' ©' 
II n II II 
a b b a 

TTT T 
z z z z 


H 






I 

1/3 


S £ to 

tfl (/) !/) 


S 


ji u a 


£ 


a 

M 


4. 

z 


^- l— l— L_ 

<2 <2 «S <£ 


u. 


«2<2 


<s 



97 



T3 

<U 

3 
C 

i 

o 
o 

00 
I 



Monthly 
Estimate, 
Method C 




Monthly 
Estimate, 
Method B 




Monthly 
Estimate, 
Method A 


T * T 
z z z 


Biweekly 
Estimate, 
Method C 




Biweekly 
Estimate, 
Method B 




Biweekly 
Estimate, 
Method A 


© _ 

T I 

z z 


Daily 
Estimate, 
Method C 




Daily 
Estimate, 
Method B 




Daily 
Estimate, 
Method A 


N=2 
OMIT 

N=9 


Network, Variance, 
Accuracy 


Strata Allocation of 
Gauges: 

Strata 1 
Strata 2 
Strata 3 



E 

3 









98 




Figure 2-3. Recorder distribution for daily measurement of precipitation at 1 1 stations in 
Okefenokee Swamp. 





99 




& VLl vs 








\ 










€^Sr '• \ 




I — ^^ • Z7 Cl= " 




I* 1 "^ A*** 




LrT ^$ J 


Figure 2-4. Recorder distribution for daily measurement of precipitation at 14 stations in 
Okefenokee Swamp. 



100 




Figure 2-5. Recorder distribution for biweekly measurement of precipitation at 1 1 
stations in Okefenokee Swamp. 



101 










Figure 2-6. Recorder distribution for biweekly measurement of precipitation at 14 
stations in Okefenokee Swamp. 



102 




Figure 2-7. Recorder distribution for monthly measurement of precipitation at 1 1 
stations in Okefenokee Swamp. 






103 




Figure 2-8. Recorder distribution for monthly measurement of precipitation at 14 
stations in Okefenokee Swamp. 



104 
monthly intervals to estimate daily average rainfall; spatial variances are nearly equal 
with the random and stratified methods (Table 2-8), indicating that over longer intervals 
spatial variability of rainfall throughout the watershed is reduced. 

Accuracy (p) for estimating daily precipitation to 3 mm with a 14-gauge network 
(a=0.20) is 0.55 (A) and 0.61 (B/C). This means there is a 30-45% probability that the 
daily precipitation will be measured within 3 mm of the true, daily volume. Bi-weekly 
and monthly estimates of daily average precipitation are more accurate (P^b/c^- 1? and 
Pa, b/c =0 - 16, respectively). The similarity of accuracies of biweekly and monthly 
estimates of daily average precipitation calculated with methods A and B/C indicate that 
little improvement is gained by stratifying the network. The network density could be 
increased by 4 recorders distributed randomly to improve accuracy of daily average 
rainfall estimated with biweekly data to 90% (p=O.10, a=0.05), and by 2 recorders 
distributed randomly to improve accuracy of daily average rainfall estimated with 
monthly data to 90% (P=0. 10, a=0.05). The network density would have to be increased 
to 1269 and 2735 stations for methods B/C and A, respectively, for the same accuracy of 
daily rainfall measurement within 3 mm, obviously an unmanageable system. Biweekly 
and monthly summaries of average daily precipitation require fewer stations to achieve 
the same level of statistical accuracy as for daily summaries. Therefore, the 14-gauge 
system should not be used to estimate rainfall day-by-day, but over longer intervals (bi- 
weekly or monthly) estimates from the network of average daily rainfall are appropriate. 
Random distribution of the gauges is sufficient, since watershed rainfall is uniform over 
the bi-weekly and monthly intervals and a minimal increase in network accuracy through 



105 
stratification would require a 10-fold increase in gauge density (Table 2-8). Little 
decrease in spatial or relative variance is apparent with stratification, although the 
current distribution among 3 strata with 2 gauges in strata 1, 5 in strata 2, and 7 in strata 
3 is appropriate for a stratified network. 

Removing 3 stations (Floyds, Sapling, and Cravens) from the network would 
reduce maintenance and management efforts, but the accuracy of rainfall estimates 
would decrease. The 3 stations selected for removal are difficult to access, but they are 
also isolated from other recorders. Relative and spatial variances of biweekly and 
monthly estimates of average daily rainfall are similar to those of the 14-gauge network 
(Table 2-8), indicating that when the data are summarized over longer intervals, 
heterogeneity in daily rainfall within the watershed is reduced. The spatial variance 
should be independent of the number of gauging stations if the watershed precipitation is 
homogeneous; an increase from 1 1 to 14 stations changes the estimate of daily spatial 
variance (Table 2-8), indicating that the differences in total variance of daily rainfall 
between the 1 1 and 14 gauge networks is partially due to the daily variability in rainfall 
in the watershed and also due to the isolation of the 1 1 gauges within the network. 
Accuracy (P) of the estimated daily precipitation measured within 3 mm with a 1 1 -gauge 
network (cc=0.20) is 0.73 (A) and 0.72 (B/C). Bi-weekly and monthly estimates of daily 
average precipitation are more accurate (P^ wc =0.20 and P^ B/c =0. 1 8, respectively). 
The similarity of accuracies of biweekly and monthly estimates of daily average 
precipitation calculated with methods A and B/C indicate that little improvement is 
gained by stratifying the network. However, great improvement occurs if the daily data 






106 
are summarized over biweekly or monthly intervals. Based on data from the 1 1 gauge 
network, a density of 20 recorders distributed randomly would be needed to improve 
accuracy of daily average rainfall estimated with biweekly data to 90% (p=0. 10, a=0.05), 
and 16 recorders distributed randomly to improve accuracy of daily average rainfall 
estimated with monthly data to 90% (p=0. 10, a=0.05). The added effort in maintaining 
the additional 3 recorders in the 14 gauge network slightly improves the biweekly and 
monthly estimates of average daily data, but they are not sufficient to provide accurate 
daily data estimates. 
Discussion of Precipitation Network Analysis 

The precipitation gauge network is intended to provide daily rainfall estimates 
throughout the swamp. This assessment of network accuracy indicates that the network 
density is sufficient to provide estimates of daily rainfall within 3 mm of actual 
precipitation volume if the daily estimates are averaged over intervals of at least 14 days. 
The accuracy of these estimates (86-87%) decreases at finer temporal resolution, because 
of the spatial variability in daily precipitation. If daily measurements are used without 
averaging over longer intervals, the accuracy in area rainfall prediction to within 3 mm is 
39-45%. A network of 1 1 stations will provide biweekly and monthly estimates of daily 
rainfall with an accuracy of 78-80%; however, the 3 stations (Sapling and Floyds 
Prairies, and Cravens Hammock) removed due to inaccessibility actually alter stratum 
delineations. Addition of 4 recorders to the existing network would permit biweekly and 
monthly estimation of daily average rainfall within 3 mm, with an accuracy of 90% 
(P=0. 10, cc=0.05). Repair of the existing gauges not used in this analysis (Sill, Gannett 



107 
Lake, Moonshine Ridge, Honey Prairie) and implementation in the recorder network 
would achieve this goal. 
Estimation of Missing Precipitation Data 

The swamp hydrology model requires bi-weekly precipitation totals for each 
500X500 m cell throughout the swamp. Swamp precipitation recorders provided a 
partial daily rainfall record, due to malfunctioning recorders; estimates of missing data 
were needed to calculate average bi-weekly precipitation for the model. The database of 
original and estimated biweekly precipitation totals were then interpolated using the 
ARCGRID KRIGING procedure (circular model) or the ARC TINNING (quintic) and 
ARC GRID procedures. The interpolated grids provided precipitation data for the 
hydrology model (see Chapter 3). 

Correlation and simple linear regression procedures were used to estimate 
missing bi-weekly totals for each recorder. Waycross, GA, data were used to estimate 
swamp rainfall coverage at recorder sites during 1930-1979. Comparisons for the best 
correlations and regressions were calculated among local recorder pairs (the nearest 1-3 
stations, using relationships in sequence of highest to lowest r^ 2 until a complete bi- 
weekly dataset resulted) for the 1980-1993 missing data. Precipitation data collected at 
Nahunta, Homerville, Folkston SW, and Waycross WSMO, GA, NOAA weather stations 
were included in these precipitation calculations. Only non-regressed, original recorder 
data were used in the correlation and regression calculations. Several regression 
relationships were necessary for each station to ensure complete data coverage during the 



108 
interval (Table 2-9). All regression pairs met assumptions of linearity, independence 
and normality of residuals, independence of data, and non-autocorrelated residuals 
(Durbin-Watson D) (Myers 1990). 
Estimation of Evapotranspiration. Inflow, and Outflow Data 

The swamp hydrology model requires estimates of biweekly, surface water 
inflow, outflow, and evapotranspiration. The outflow points for surface water included 
in the swamp hydrology model were the Suwannee River, Cypress Creek, and 
Sweetwater Creek, near Fargo, GA, and the St. Marys River near Moniac, GA (Figure 2- 
9). Data retrieved from USGS gauges that measured daily Suwannee and St. Marys 
Rivers flow rates for 1930-1993 provided biweekly estimates of surface outflow volume 
for the hydrology model and analysis of the sill's effects on swamp hydrology. The St. 
Marys River gauge at Moniac, GA, was dismantled in 1989 and reinstalled in 1991 ; 
missing data for this station for 1989-1991 were estimated with regressions with the St. 
Marys, MacClenney, FL, USGS flow gauge, or the Suwannee River, Fargo, GA, flow 
gauge (Table 2-10). Missing Suwannee River flow data were estimated from regression 
relationships with the St. Marys River gauges at Moniac, GA, and Macclenney, FL. Flow 
gauges were not installed at Cypress and Sweetwater Creeks. Measurements of biweekly 
flow volume at these stations were estimated from regression relationships established 
between the creek water depth and Suwannee River flow recorded daily during 1991- 
1993 (Table 2-10). Only non-regressed, original recorder data were used in the 



109 



Table 2-9. Best correlation pairs and regression equations used to estimate missing 
precipitation recorder data for use in HYDRO-MODEL, during 1930-1993. 



Interval and 

Predicted 

Station (Y) 


Predictor Station 
(X) 


r 2 ■ 
r ld j 


Regression Equation to 
Estimate Y 


1930-1947 








Sapp Prairie 
(digital) 


Waycross 


0.2775 


y = 0.617487x + 0.491895 


Craven's 
Hammock 


Waycross 


0.3632 


y = 0.469907x4- 2.585811 


Durdin Prairie 


Waycross 


0.2439 


y = 2.31721x 1/2 - 0.141917 


Coffee Bay 


Waycross 


0.3805 


y = 0.508594x+ 1.15652 


1930-1948 








SCRA 


Waycross 


0.3849 


y = 0.657773x + 2.176066 


Folkston SW 


Waycross 


0.4478 


y = 0.699435x+ 1.629737 


Moonshine Ridge 


Waycross 


0.2527 


y = 0.630105x+ 1.239003 


1930-1956 








Nahunta 


Waycross 


0.5238 


y = 0.816304x+ 1.122031 


Homerville 


Waycross 


0.5676 


y = 0.806905x4- 1.197628 


1930-1978 








Waycross WSMO 


Waycross 


0.7091 


y 1/2 = 0.1 8902x4- 1.002779 


1930-1979 








Soldier's Camp 


Waycross 


0.2629 


y = 0.963607x 4- 0.775384 


Suwannee Creek 
(chart) 


Waycross 


0.1818 


y=1.207899x + 0.101984 


Territory Prairie 


Waycross 


0.2374 


y = 0.488879x+ 1.230789 


Suwannee River 


Waycross 


0.5141 


y = 0.56549x4- 2. 114985 


Floyd's Prairie 


Waycross 


0.4429 


y = 0.598386x+ 1.885058 


Honey Prairie 


Waycross 


0.4006 


y = -0.2606X + 4.381847 



Table 2-9~continued. 



110 



Interval and 

Predicted 

Station (Y) 


Predictor Station 
(X) 


r 2 ■ 

1 adj 


Regression Equation to 
Estimate Y 


1930-1980 








SCFSP 


Waycross 


0.4609 


y = 0.69817x+ 1.678838 


1948-1979 








Coffee Bay 


SCRA 


0.5761 


y = 0.6445 15x + 0.257361 


Durdin Prairie 


Folkston 


0.7483 


y = 0.6527x + 0.796215 


Cravens 
Hammock 


SCRA 


0.5813 


y = 0.747346x + 0.449443 


1948-1981 








Moonshine Ridge 


SCRA 


0.3452 


y = 0.6367 17x + 0.691347 


1948-1993 








Sapp Prairie 
(digital) 


SCRA 


0.9185 


y = 0.92291x- 0.02865 


1979-1991 








Coffee Bay 


SCFSP 


0.6228 


y = 0.751514x + 0.126923 


Honey Prairie 


Seagrove Lake 


0.7839 


y=1.188659x- 0.430333 


Durdin Prairie 


Territory Prairie 


0.8406 


y = 0.785827x+ 1.6441 13 


Suwannee River 


SCFSP 


0.8498 


y = 0.84687x + 0.233137 


Floyd's Prairie 


SCFSP 


0.7686 


y - 0.938372x + 0.279866 


Sapp Prairie 
(digital) 


Seagrove Lake 


0.5791 


y = 0.568588x+ 1.84186 


Sapling Prairie 


Chase Prairie 


0.2712 


y = 0.435358x + 2.689235 


Craven's 
Hammock 


SCFSP 


0.4296 


y = 0.6843 17x+ 1.287402 


1980-1982 








Kingfisher 
Landing 


Suwannee Creek 
(chart) 


0.4003 


y = 0.446332x + 2.128122 



Table 2-9-continued. 



Ill 



Interval and 

Predicted 

Station (Y) 


Predictor Station 
(X) 


r 2 * 

1 adj 


Regression Equation to 
Estimate Y 


Double Lakes 


Suwannee Creek 
(chart) 


0.6166 


y = 0.61941x + 1.128997 


Sill (Brown Trail) 


Suwannee Creek 
(chart) 


0.5880 


y = 0.44801x + 2.326436 


Sapp Prairie 
(chart) 


Soldier's Camp 


0.2685 


y = 0.504175x + 0.626725 


1980-1993 








SCFSP 


Chase Prairie 


0.4123 


y = 0.715778x + 2.776936 


Territory Prairie 


Chase Prairie 


0.5189 


y = 0.862786x + 0.831206 


Chase Prairie 


SCFSP 


0.4123 


y = 0.581559x + 0.441311 


Double Lakes 


Chase Prairie 


0.4150 


y = 0.59181 lx+ 1.345497 


Kingfisher 
Landing 


Double Lakes 


0.3453 


y = 0.618564x+ 1.263495 


Seagrove Lake 


Territory Prairie 


0.3489 


y = 0.537173x + 0.605698 


Sill (Brown Trail) 


Soldier's Camp 


0.5443 


y = 0.590046x + 0.394465 


Soldier's Camp 


Sill (Brown Trail) 
Sapp Prairie (chart) 
Seagrove Lake 


0.5443 
0.2685 
0.2581 


y - 0.967946x + 2.008823 
y = 0.577914x + 2.976061 
y = 0.502014x + 2.757349 


Suwannee Creek 
(chart) 


Double Lakes 
Sill (Brown Trail) 


0.6266 
0.5880 


y=1.00928x + 0.384732 
y=1.34424x- 1.249998 


Moonshine Ridge 


SCFSP 


0.4274 


y - 0.776344x + 0.62422 


1982-1993 








Gannett Lake 


SCFSP 


0.4271 


y = 0.57035x + 0.313958 


Sill (Brown Trail) 


SCFSP 


0.4680 


y = 0.716572x- 0.29646 


Sapp Prairie 
(chart) 


Sill (Brown Trail) 


0.2149 


y = 0.519188x + 0.261563 


1992-1993 












Table 2-9-continued. 



112 



Interval and 


Predictor Station 




Regression Equation to 


Predicted 


(X) 


r 2 ■ 

r «dj 


Estimate Y 


Station (Y) 








SCFSP 


Suwannee River 


0.8498 


y = 0.954722x + 0.585187 




Floyd's Prairie 


0.7686 


y = 0.938372x + 0.27866 


Durdin Prairie 


Territory Prairie 


0.8406 


y= 1. 0742 18x- 0.967435 




Chase Prairie 


0.8346 


y = 0.872464x - 0.712295 




Double Lakes 


0.4802 


y = 0.658392x + 0.355303 




Kingfisher Landing 


0.5398 


y = 0.951845x- 0.520653 




Gannett Lake 


0.1848 


y = 0.246431x + 0.440111 


Honey Prairie 


Seagrove Lake 


0.7839 


y=1.188659x- 0.430333 


Sapp Prairie 


Suwannee River 


0.7424 


y = 0.822689x4- 0.610583 


(digital) 


Craven's Hammock 


0.7411 


y = 0.944445x 4- 0.339577 




Floyd's Prairie 


0.7312 


y = 0.646029x4- 0.819186 




Honey Prairie 


0.6868 


y = 0.872658x + 0.99355 




Coffee Bay 


0.6544 


y = 0.776927x - 0.036708 




Seagrove Lake 


0.5791 


y=1.045958x- 0.536715 


Suwannee River 


Sapp Prairie (digital) 


0.7424 


y = 0.822689x4- 0.610583 




Floyd's Prairie 


0.7026 


y= 1.027007x4- 1.139322 




Sill (Brown Trail) 


0.4215 


y = 0.727799x4- 0.724884 


Territory Prairie 


Durdin Prairie 


0.8406 


y = 0.785827x4- 1.644113 


Floyd's Prairie 


Durdin Prairie 


0.7536 


y = 0.768902x + 0.674328 


Craven's 


Coffee Bay 


0.7023 


y = 0.777602x4- 0.294183 


Hammock 


Sapling Prairie 


0.5828 


y = 0.608385x 4- 1.188687 


SCRA 


Craven's Hammock 


0.5813 


y = 0.747346x 4- 0.449443 


Coffee Bay 


Craven's Hammock 


0.7023 


y = 0.9 1 02 15x+ 1.379420 




Sapling Prairie 


0.5828 


y = 0.777694x + 0.1 89880 


Seagrove Lake 


Honey Prairie 


0.7839 


y=1.188659x- 0.430333 



8 All regression relationships were significant at P < 0.05. 



114 



■ 

CN 

I 





u 










u 








<u 










<D 






ItM 


v> 






£ 




cfl 






o 


fi 








w 






1? 


a 






00 




£ 






© II 


rn 






ci 




00 








os 






m 




00 






© 






V 




— " 








V 

o 

CN 






fa 

CN 

00 




V 
o 

sa 

CN 




O 


a 


CN 






<n 




r~- 






# o 


CN 






rj- 




oo 




1 5 

-2 1> 


<B 


m 






00 




00 




03 


o 






rn 




CN 




S 


o 

+ 






+ 




© 

+ 




e 
,© 








~x 




Os 




1> Sfl 


rn 






VO 




© 






"5 
on 

c 


CN 






o 

Os 








O 






<n 




«r> 




g U 


en 


CN 






VO 




^ 




£ £^ 


J 


O 






O 




o 




00 ^ 


II 






© 




II 




fN 




>> 






II 

>s 




>> 




© 




^ — 










_ 








o 






o 




o 






0- 


o 
o 

o 






o 
© 

o 
m 




o 
o 

© 

CN 






*5* 


OS 






en 




SO 






■ 


O 






o 




r~ 






■u 


00 

© 






oo 
©' 




© 




Li 


Predicted 
Measure 


00 


o 

a 

'S 
o 

2 


o 


en 


fa 

00 


c/i 

1) 
E 


O 


i 

1 ^ 
=5 

CO tu 


ctor 
ure 




M 




c 
c 


a 

> 

o 

E 




E 




1/3 E 


Predi 
Meas 


C/3 


u 

2 


15 


1 

oo 


o 

60 

t- 
A 


i 

3 
CO 


0" 

DO 

s- 


U U 


















M 




0k 














% 


la 

B 

> 

a 


0) 

> 






J! 
0) 
V 

•- 




k. 












u 




u 




4> 
OS 


o 




•SS 








if. 
8 




£ 


1 


7 


"5 

O 

- 




L- 

a 




o. 






E 

u 










U 




it 



115 
regression calculations. All regression pairs met assumptions of linearity, independence 
and normality of residuals, independence of data, and non-autocorrelated residuals. 

The hydrology model required biweekly surface water inflow from creeks along 
the swamp perimeter. Creek flow into the swamp is significant only along the northwest 
boundary, accounting for an estimated 20% of the swamp annual water budget (Blood 
1981, Rykiel 1977). Water depths were measured on permanently installed staff gauges 
in 7 creeks (Bear Branch, Cane Creek, Gum Swamp, Suwannee Creek, Greasy Branch, 
Surveyor's Creek, Black River) every 4-6 weeks during 1991-1995 to establish water 
depth relationships with the Suwannee River (Figure 2-9). Creek flow rate was also 
measured with a General Oceanics, Inc., flow gauge, converted to flow volume based on 
creek dimensions measured at the recording station, and regressed with concurrently 
collected staff data to relate creek water depth to estimated creek flow volume (Table 2- 
11). Regression relationships between creek water depth and Suwannee River flow were 
used to extend the creek water depth estimates back to 1941 (Table 2-12); the estimated 
creek water depths were then converted to estimated creek flow volumes using these 
regression relationships. All regression pairs met assumptions of linearity, 
independence and normality of residuals, independence of data, and non-autocorrelated 
residuals. 

Approximately 80% of the water that leaves the swamp does so through 
evapotranspiration, or ET (Yin and Brook 1992a, Yin 1990, Hyatt 1984, Blood 1981, 
Rykiel 1977). This parameter was not measured directly in this study but was estimated 
for the hydrology model database using Thorthwaite's equation for monthly potential 



116 



1 

M 

o 

E 

X/i 

<U 

•4-* 

1 

co 
<U 



Q. 
I 



oo 
9 

'55 

3 

& 

1 



1 

o 

I 

i 

1 

O 
J 

o 

8 

an 

c 

I 

1 



U 

1 



•*> 
















e 


Tj- 


On 


*-* 


<**N. 


«^> 


r- 


On 


c m 


«/"> 


r^ 


m 


oo 


On 


>— ' 


(N 


o © 


O 


o 


o 


© 


C> 


i— « 


— i 


2 ii 


VI 


VI 


VI 


VI 


VI 


VI 


VI 


B 1 


53 


53 


53 


53 


53 


53 


53 


5 e 


2 


■4— > 


S 


2 


s 




B 


U to 


oo 

00 
00 


00 


00 


oo 


oo 


00 

NO 


V) 




ir> 

00 


f» 




00 




so 






r*-> 


NO 


so 


CnI 


r<^ 


^"* 




m 


in 


</-> 


Tj- 


O 


On 




. 


ON 


rr 


o 


CN) 


SO 


w-> 


c 


<N 


00 


m 


o 


m 


cr> 


m* 




(N 
i 


00 


On 


Tf' 


<n 




On 


■ 


C« 


■ 


|M| 


1 


so 


i 


rj 


s 


~x 


X 


1 


X 


i 


X 


i 


P-H 


>T 


5 


<N 


X 


r^ 


X 


CM 


so 


T— H 


<N 


On 


00 


H 


ON 


SO 


00 


SO 


* 


O 


NO 


© 


o 


NO 


rf 


On 


(N 


SO 


»/-> 




5 


l"» 


ir> 


m 


O 


00 


00 


en 


TT 


v—4 


</-> 


■<1- 


* 


«r> 


m 


B 


p^ 


^ 


00 


sb 


NO 


^ 


OS 


& 


* 


>— I 


rf' 




NO 


<N 


»^ 


II 


11 


II 


II 


II 


II 


II 


ec 


>. 


>. 


>, 


>> 


>. 


>> 


>-, 




ir> 


m 


_ 


so 


Wi 




so 


- 


(N 


-* 


O 


© 


»/"> 


O 


On 




^H 


SO 


O 


<N 


m 


O 


sO 




O 


«— 


o 


o 


o 


o 


O 




© 


© 


o 


© 


o 


©' 


O 




<N 


© 


3 


r- 


^ - 


« 


O 


T. 


</-> 


NO 


oo 


00 


<n 


in 


r- 


PN 


*fr 


m 


C5 


00 


IN 




00 


rv| 


00 


O 


m 


00 


SO 




o 


©' 


©' 


© 


o 


o 


d 


"9. U 


£ 


£ 


£ 


£ 


^ 


^ 


£ 


fi C 


o 


O 


o 


O 


o 


o 


o 


£ 5 
















S 


u- 


u* 


E 


u. 


b 


b 


iQ 98 


4< 


.M 


41 


4* 


M 


41 


4d 


cy o 





s 





6 


6 


B 


B 


£S 


u 


p 


<u 


P 


p 


p 


B 


E 


E 


E 


E 


E 


E 


E 


u 


u 


o 


u 


u 


u 





»- <L> 


CJ 














licto 


53 


53 


53 


53 


53 


53 


53 


a 

40 


CO 


2 


3 


C8 

■4-J 


2 


a 


cy o 

ft. * 


00 


OO 


oo 


00 


00 


00 


OO 


41 


41 


4J 


M 


41 


41 


M 





B 


6 


B 





<u 


6 


^^ 


p 


ID 


p 


p 


p 


B 


B 




i— 


E 


E 


E 


E 


E 


E 










u 


U 





U 

















S 


L 




^ 






>.-= 


a 


a 


o 




6 

■- e 

08 08 


Black 
River 


S 8 


08 B 


El 


E | 


fa 


Cv u 


08 u 


- 1 - 


3 i 


3 - 


c 3 . U 


u 


efi qq 





e> « 


oo U 






117 



Table 2-12. Regression relationships used to estimate water depths at staffs in 
northwestern creeks from flow measurements at the Suwannee River-Fargo gauge. 



Creek 


Predictor 
Measure 


Predicted 
Measure 


r 2 


P 


Regression Equation 


Bear 
Branch 


River 
Flow 


Creek Staff 


0.3484 


0.0552 


y = 0.005794x + 0.291364 


Cane 
Creek 


River 
Flow 


Creek Staff 


0.6773 


0.0001 


y = 0.01616x + 0.085627 


Gum 
Swamp 


River 
Flow 


Creek 
Staff' 2 


0.4739 


0.0019 


y = 0.074380x" 2 + 0.546947 


Suwannee 
Creek 


River 
Flow 


Creek 
Staff" 2 


0.7270 


0.0001 


y = 0.111256x ,7 + 0.645043 


Greasy 
Branch 


River 
Flow 


Creek 
Staff 1 ' 2 


0.7187 


0.0001 


y = 0.88629x la + 0.063674 


Surveyor's 
Creek 


River 
Flow 


Creek 
Staff 4 ' 2 


0.7434 


0.0001 


y = 0.126597x 1/2 + 0.409234 


Black 
River 


River 
Flow 


Creek 
Staff" 2 


0.7875 


0.0001 


y - 0.076624x 1/2 + 0.202671 












118 

evapotranspiration (Thornthwaite 1948). The relationship creates a ratio of mean 

monthly air temperature and heat index as follows: 

PE = (1.62)b(10T/I)\ 

where 

PE = monthly potential evapotranspiration (cm) 

b = monthly latitude coefficient to account for seasonal radiation (Table 2-13) 

T= monthly average daily temperature °C 

a = 67.5 x 10" 8 I 3 - 77.1 x 10" 6 ! 2 + 0.01791 + 0.492 

12 

I = heat index = fttJSf* 1 , where m=monthly periods, t=mean monthly air temperature, 

m=l 
°C. 



Table 2-13. Monthly latitude adjustment to account for seasonal radiation in calculation 
of Thornthwaite's PE (from Thornthwaite (1948)). 



January 


February 


March 


April 


May 


June 


0.90 


0.87 


1.03 


1.08 


1.18 


1.17 


July 


August 


September 


October 


November 


December 


1.20 


1.14 


1.03 


0.98 


0.89 


0.88 



Daily air temperatures were recorded for various intervals at 6 NOAA weather stations 
around the swamp (Figure 2-9). Regression relationships between these stations were 
used to estimate missing daily average temperature, which were used in the estimate of 
PE (Table 2-14). Mitsch and Gosselink (1986) suggest that in wetland environments, 
potential evapotranspiration is nearly equivalent to actual evapotranspiration since water 
availability is rarely limited. M. Focazio (USGS, unpublished data) estimated that 



119 



Table 2-14. Regression equations used to estimate missing daily maximum air 
temperature at NOAA weather stations around Okefenokee National Wildlife Refuge. 



Predicted 


Data 


Predictor 


r 2 ' 


Regression Equation 


Station 


Interval 


Station 






Homerville 


1930-1955 


W4NE 


0.9740 


y = 0.986729x + 0.173426 




1956-1993 


Folkston 


0.9882 


y = 0.918883x- 1.808971 


Folkston 


1930-1947 


W4NE 


0.9240 


y= 1.09373 lx- 0.887017 




1948-1993 


Homerville 


0.9882 


y=1.075527x- 3.057219 


Fargo 


1930-1981 


W4NE 


0.9173 


y=1.07418x- 0.831709 


(SCFSP) 


1982-1993 


Folkston 


0.9735 


y = 0.960154x- 2.714077 


Nahunta 


1930-1955 


VV4NE 


0.9691 


y - 0.967064x + 0.409845 




1956-1993 


Folkston 


0.9679 


y - 0. 5675 15x+ 16. 866641 


WSMO 


1930-1978 


W4NE 


0.9180 


y=1.0968x- 1.034505 




1979-1993 


Homerville 


0.9820 


y=1.01434x- 1.202566 


W4NE 


1930-1993 


Homerville 


0.9793 


y=1.10042x- 2.289395 



All regression relationships are significant at P = 0.0001. 



Thomthwaite's PE underestimates actual evapotranspiration in cattail (Typha spp.) 
swamp up to 37%; a comparable adjustment to the calculated ET values for Okefenokee 
Swamp was made in the hydrology model (see Chapter 3) to refine model output. Yin 
and Brook (1992a) also found Thomthwaite's PE to be well-correlated with actual 
evapotranspiration rates in the swamp. The monthly ET volume was halved to provide 
biweekly volumes for the hydrology model. Biweekly estimates were interpolated 
among recorder stations using ARCINFO's tinning (quintic) procedure to create 
biweekly ET surfaces, and gridded at 500x500 m cell resolution for use in the hydrology 
model. 



120 



Swamp Basin Delineation and Characterization 



Water level data recorded at gauges during 1980-1995 illustrate the spatial 
connectivity as well as regional variabilities of the Okefenokee Swamp hydrologic 
environment (Table 2-2). Highest water elevations were recorded in the North and 
Northeast (Double Lakes, Kingfisher Landing, Durdin Prairie, Sapling Prairie), where 
peat surface elevations are highest, and in Honey Prairie, where a northwest to southeast 
peat surface ridge runs between Honey and Blackjack Islands. Lowest water surface 
elevations were recorded in the Southwest drainages (Suwannee River, Sill, Sweetwater 
Creek, Cypress Creek) and St. Mary's River basin (Soldiers Camp). Greatest variability 
in water surface elevation occurred in high flow areas, such as the creeks and tributaries 
to the Suwannee and St. Mary's Rivers (Figure 2-10). During 1992-1995 water surface 
elevations at the Sill, Suwannee River, Craven's Hammock, Cypress Creek, and 
Suwannee Creek changed 2. 16-1.06 m. Greatest changes in water surface elevations in a 
day were recorded at Suwannee, Sweetwater, and Cypress Creeks, the Sill, Craven's 
Hammock, and Territory Prairie (+0.40 - +0.29 m). All of these stations are located in 
areas of channelized flow. Territory Prairie experiences a drop towards Chase Prairie of 
0.6 m in peat surface elevation in the area around the recorder. The change in elevation 
localizes the area's water flow into the maintained canoe trail near the recorder station. 
Prairies, lakes, and canals had the smallest high to low water level ranges. Maximum 
water surface elevation changes in Chase, Durdin, and Honey Prairies, Double Lakes, 
Moonshine Ridge ditch, and Kingfisher Landing canal ranged 0.31- 0.45 m; daily 



121 



O 



■8 



JO 



CD W W 

c75 q: J2 w 







On 

ON 

I 

(N 

On 

On 

&d 
c 
•g 

3 

a 
£ 

00 

u 

<D 

J* 

o 

c 

^4-1 

o 

1) 
£ 



on 
C 

_o 

■♦-» 
aj 

o 

o 



A 



03 

•o 

-o 
t> 

T3 

I— 
O 
o 



Q. 



to 

s- 
3 
O 

c 
o 

O 

O 
o 

c 

■g 

> 

o 

I 

Cn| 

I 

E 



122 
changes were generally less than a centimeter. There are 5 "basins" represented by the 
spatial variability of the swamp hydrology (Figure 2-11, Table 2-2). Each of these areas 
follows the overall seasonal trends in water surface elevation, but the magnitude of these 
trends varies among the basins (Figure 2-12). Greatest seasonal and annual variability in 
water surface elevation occurs in the northwestern region; water surface elevation in this 
area is probably controlled by seasonal rainfall, primarily because much of the water is 
contributed by streams in the watershed to the west of the swamp. In contrast the least 
seasonal and annual variability in water surface elevation occurs in the Northeast. This 
may be due to groundwater inflows or restricted outflow creating a perched water 
surface. The central region has intermediate variability. Most of the water in this region 
is contributed by precipitation, and water surface elevation declines rapidly during 
periods of high evaporative demand. There may be some groundwater exchange in this 
area through springs, although this component of the water budget may be relatively 
minor, and probably originates in the surficial aquifer (Rykiel 1984, 1977, Patten and 
Matis 1984, 1982). The Southeast and Southwest basins are somewhat hydrologically 
isolated from the rest of the swamp by a surface ridge created by large islands 
(Blackjack, Mitchell, Soldiers Camp, Honey, Billy, Pocket). The southwest basin 
contributes to the Suwannee River outside of the refuge boundary, and the Southeast 
basin forms the headwaters of the St. Mary's River. These areas show intermediate 
fluctuations of the central region, and variability like the northwestern region at the basin 
low points (Soldiers Camp in the Southeast and Cypress and Sweetwater Creeks in the 
Southwest). The role of precipitation, evapotranspiration, inflows, and outflows in 









123 












T~*~^ \ Northwest S # If Northeast 






\ 


1 J \, 














Central & 








^ • 7 






• 


W - " p 






w» 








• 

> rj Southwest 


1 ^r7~w7 

1 A ^Southeast 

6^^^ / 




Figure 2-11. Water level recorder locations and hydrologic basins in Okefenokee 
Swamp. 





124 



2.0 
1.5 - 
1.0 - 
0.5 - 



0.0 
2.0 



Central Basin 
(Chesser Prairie) 




1.5 
1.0 
0.5 - 



0.0 
2.0 



Southwest 
(Cypress Creek) 




a 
a 

| 

CO 



CD 
O) 

a 

0) 
> 
< 



1.5 - 

1.0 

0.5 



0.0 
2.0 



Northeast 
(Durdin Prairie) 




1.5 - 
1.0 - 
0.5 - 



Southeast 
(Moonshine Ridge) 




Figure 2-12. Trends in water level fluctuations in the Okefenokee Swamp hydrologic 
basins. 



125 
controlling swamp water depth varies with the basin. These relationships are explored in 
the hydrology model discussion in Chapter 3. 

Approximately 80% of the swamp water budget is contributed by precipitation 
and removed by evapotranspiration (Yin 1990, Hyatt 1984, Blood 1981, Rykiel 1977). 
The effects of these processes on swamp water level vary seasonally. Evapotranspiration 
demands are unimodal, with a peak during May-August (Figure 2-13). Precipitation 
peaks during June-September and again during January-March (Figure 2-14). The higher 
precipitation volume during June-September does not usually result in high water levels 
because of the evapotranspiration demand; water levels are more likely to rise with the 
increased precipitation volume in January-March, when evapotranspiration demands are 
lowest. Evapotranspiration rates are not uniform across the swamp, but reflect 
differences in vegetation composition; evapotranspiration has a greater effect on swamp 
water level fluctuations in the eastern swamp than in the west (see Chapter 3). River 
outflows and creek inflows account for approximately 10-30% of the overall swamp 
water budget (Blood 1981, Rykiel 1977). Fluctuations in inflows and outflows follow 
those of precipitation, with biannual peaks in February-April and August-October 
(Figures 2-15 and 2-16). Water entering the swamp via creeks and rivers impacts the 
western swamp, although minimal surficial input occurs from streams along the eastern 
perimeter (Brook and Hyatt 1985, Hyatt 1984, Hyatt and Brook 1984, Rykiel 1984, 
1977)(also see Chapter 3). Groundwater exchange is estimated at 3-5% of the swamp 
water budget; the sources, variability, and extent of this component are unknown (Brook 
and Hyatt 1985, Hyatt 1984, Hyatt and Brook 1984, Rykiel 1984, 1977). 



126 



■ • : ' ■ ' ~ ' — 



i 



o 

h- _> 
Q 




...:... :. , ___. ___ , , .:£ ; y. 



_i .„ ,,,.'„ :-.... •..-... : ,, ■-•. .,, 



a 
a> 
eo 

CO 

3 
< 



>> 
I - 3 



' C 

0) o 

i ^ 

"3 



| ' " — : : — —' — ; 



CO 



~ . I . — ■ , — _______ 



_ CO 



i r 



u 

a 
< 



_3 
0) 



c 

CO 



OOCDTfOJOODCOTj-CMO 

(iuo) uoftBJidsuBijodBAg A|ifluo|/\| eBejeAV 



ON 



i 



c. 



I 

— ■ 

_> 
__ 

O 

_) 

_c 



1 

_3 
<D 

c 
o 

1 

c 
2 

& 
> 
_> 

u 

-0 

2 

1 



1 

o 



r*i 



& 

3 
60 



127 



o o 

CO CO 



j ' — 



" ■■.■■■■ ■ <■■ 

■ . • 



■■" ■ — ■---;-■■- .— —T-— ■■■■■' ■" 

..■■■■■.,.■■■.,■■.,■■■■■ ■ ■■■■■■• ■ ••...■■■■' .■....•.■•.,,. ■■■■■ 



■ ■ ■■■ ■■••■• 



^\a;^\«^ 



- 



■ — ■ , 



• " ••■ • ■ ■■■■ " ' ■' aw . . .. „ . , ■■,. 



— , — , — __, — 



i 



C 



1 "■ lll : . ' ■ ■:'■■■■■ ; ■■■ ■ 

_ — ,,,, , ,, 



I I I 1 1 1 1— 

COKCOIO^COCMt- 

(oio) uoaejjdpajd A|muo|fl a6ej8AV 



© 

o 



> 
o 

z 



u 
O 



a 

CO 

0) 

3 
< 



3 
-i 



3 5 



(0 



a 
< 



(0 



0) 



c 

TO 

-5 



o\ 

ON 

■ 

o 
m 

i 



09 

3 



j 

o 

I 

.E 
2 



C8 
l 



§ 
1 

a 

I 

>> 
2 
e 

1 

I 

I 



P 



128 



CO 

Is 

a 

S. 4 

I 
1 3- 



Q 2 
e 

D) 
(0 



0) 
> 
< 



1 ~ 



150 

% 125 

I 

n 
K 100 

1 

I 75- 



(5 

Q 50 

a 

to 

fc 25 

> 
< 



I 



Black River 



^M Cane Creek 

I 1 



I Greasy Branch 

Gum Slough 
1 Surveyor's Creek 
] Suwannee Creek 



JUL 








Bear Branch 


HP! 



\ \ \ I ''T" "7~ r T~ l I 1 " 1 ""i 1111 

Jan Feb Mar Apr May June July Aug Sept Oct Nov Dec 

Month 



Figure 2-15. Average daily inflow estimated for northwestern creeks entering 
Okefenokee Swamp during 1930-1993. 



129 



S 



(suio) 3}ey AAOy^no A|;e(] a6eJ9AV J3Aiy aauueMng 

9 8 5 




o 

1 

© 

CO 



e 

£ 
o 



I 



<X> 

> 



CO 



> 
cr 

C9 

c 
c 

I 

3 
CO 

D 



00 (0 ^ CM 

(suio) s;ey MO^no a I! b Q eBejdAV 






u 

o 

— 

3 
O 



cs 
o 

2 

o 



g 

3 
(50 

E 



130 
Effects of the Suwannee River Sill on Swamp Water Level Conditions 

Changes in water surface elevation that have occurred at SCRA and SCFSP 
during 1941-1995 suggest effects of the Suwannee River sill on swamp hydrology; 
recorder data from these stations provide an indication of the sill's effects independent of 
results from the swamp hydrology model. The swamp hydrology model presents a more 
complete picture of the sill's spatial effects; however, the data collected at the SCRA and 
SCFSP gauges are the only original data available from the pre-sill period. Pre-sill 
starting water depths for the swamp hydrology model are based on SCFSP and SCRA 
gauges and their regression relationships with other recorders, under with-sill conditions. 
The effects of the sill discussed here are calculated from pre- and with-sill data collected 
only at the SCRA and SCFSP gauges. Effects estimated by the swamp hydrology model 
at other recorder stations are discussed in Chapter 3. Comparisons between pre- and 
with-sill intervals are with t-tests; variances are compared with F-tests. Comparisons 
among decades are with analysis of variance and Tukey's test for differences among 
means. Flow and precipitation data were normalized with log transformations. 

The sill has affected the swamp hydrologic environment, although its effects vary 
with distance from the structure (Chapter 3). Although Yin and Brook (1992b) and Yin 
(1990) reported that discharge volume from the Suwannee River decreased and St. Marys 
flow variability increased after sill construction, their results may have reflected a data 
record that did not include a recent period of low rainfall, and insufficient topographic 
information (see topography map development and discussion, this chapter). An 



131 
additional 7 years (1987-1993) of with-sill Suwannee River and St. Marys River flow 
data recorded at Fargo, GA, and Moniac, FL, respectively, were included in my analyses. 
Log-transformed, biweekly total flows measured at these stations during pre-sill (1930- 
1959) and with-sill (1960-1993) intervals show increased biweekly flow volume at both 
stations in the with-sill period (P=0.0001) (Table 2-15). Extension of the flow record to 
1993 also suggests that variability of the Suwannee River flow decreased during the 
with-sill interval (PO.0001), whereas variability of the St. Marys River flow did not 
change with installation of the sill (P=0. 1452). Yin and Brook (1992b) and Yin (1990) 
attributed the increased flow volume and decreased flow variability recorded at the St. 
Marys River to the sill; they hypothesized that the impounding effect of the sill was 
causing these changes. The changes in flow volume and variability at the St. Marys and 
Suwannee River gauges indicated in this study most likely reflect concurrent changes in 
rainfall patterns, not sill-induced modifications of water flow within the swamp. 
Although t-test comparisons of log-transformed precipitation volumes recorded at SCFSP 
and SCRA showed no differences in pre- and with-sill biweekly totals, variability of 
rainfall volume was slightly higher at the SCRA recorder following sill construction, and 
January precipitation totals were higher and September totals lower at both stations 
during the with-sill period (Table 2-16). Suwannee and St. Marys River flow volumes 
were also higher in January following sill construction (Table 2-15). These higher flows 
may have resulted in part from changes in air temperatures and evaporative demands; 
pre-sill estimated evapotranspiration volumes were higher at most NOAA weather 
stations in the watershed during all months except May, July, September, and November 



132 



00 



I 

t . 

2 5 

D Os 

5 <n 

S On 



- -O 

M __'" 

"O ~ 

<D 00 

hi 

3 >- 

a > 

'Z> 

2 | 

1 I 

o 3 

a « 

Cm 12 

o o 

§ .2 

D. « 

cn o 



CN 



y 

H co 



•r 5 u. 

* "l A 



c 

O on 

J = 

E « — 

S. >S A 



IIP 



Si 



m oo r^i o 

(NO O CN 
O CN O O 



— SO 00 00 

or- oo cn 

o o r- cn 

o O — ' — 



v© </-) 

cn CM 

00 00 

CN — 



or- so cn 

O cn oo cn 

cn cn m o 

n (N Os — 



oo oo oo oo oo oo oo 



o •*! 

° o 



"* * — eN 

o os o o 

O SO O O 

ot o o 

o o o o 



CN cn 

o © 
o o 

© o 



SO 00 V~l •— 

cn cn cn © 

CN so »- F> 

— — — < o 

o © © © 



os o r» o 

SO Tf — ' 00 

(NO — m 

© o o © 



so 
cn 



o 






© 

S .2 .-j. 

S J* 1 
Be > «, 

i cj 

£ « 



5 |i 

is? 



•- c I 

OO W U 



c 
o 



09 

55 






cn CN 



r- cn 

Os *■* 



Z4 i-J "* «» 



l-< cn 
cn IS 

"t" rn 



cn Ti- 
en os 
ci cn 



o 



oo 

Os 



m 



CN 



Os" rn g 



i-i oS 



SO 



Cn 



Os 

en 



O so 
CN T 



cn os 



CN os 



— — so r-. 
cn m — <— 



CN 
CN 



2 <N 

. m 
o cn 

— CN 



00 



r- 

os 



CN 



cn 

so 



CN SO 

m os 



•O CN 

rn cn 



°°o £3 






Os 



CN 



<N 12 



3* 



CN 
CN 



m 
m 



CM 



os ^ 

TT CN 



o J2 s « 

so cn oo 



d „ 



O SO 



SO T 

— 00 

p-i i Os 



OJ 



Os 



CN 



r*- 



m 
r- 



<n 



Os 



— ~ 'st v : r>- l -. cn 

^ cn _; m „• o _; os 



oo o 

m •-< 

o K 



Tj" OS 

O rn 



2 " 

Q 






«J 3 

&o oo 



> 


1 


B 
> 


CJ 

> 


u 

<u 

> 


> 


i- 

> 


> 


CCj 


cd, 


2 


2 


2 


& 


2 


2 


V5 


c 
c 


co 

CO 




C/3 


(U 


CA 

& 

CO 


u 

1) 
B 
3 


2 


t 


s 


CO 


2 


e0 


2 


5 


■*-• 


3 


♦-* 


3 


■«--' 


3 




3 


ai 


oo 


00 


00 


00 


OO 


00 


00 



> 


> 


u. 
> 


> 


2 


ft5 


2 


2 


CO 


<u 

c 


CO 


u 
Q 

c 


CO 


c 


CO 


& 


S 


t 


2 


I 


■*-» 


3 




9 


00 


00 


00 


O0 



m 

Os 
Os 

i 

o 
m 
os 



c 

CO 



CO 

2 



o. 



CO 

2 



4> 



r- _ 



os 2t» 



oo oo 

so rn 

cn o 

— CN 



O SO 
Cj cn 

l-N OS 



O 
cn 

— i S 






o 



CJ 



2 S 

£■§ 

CO c 

«J 3 

oo oo 



3 



133 



•o 

3 

S 

'-*■* 
c 
o 
t> 

I 
I 



CN 

U 

•s 



rison 
ances 
F 


so m 


IO <— 


ON O 


CN CN 


h« <r> 


fN </■> 


t*« r^ 


SO o 


o </i 


rr so 


«•; a 


<* so 


oo >n 


— < *-* 


"* o 


00 o 




— M 


>/-> 00 


r- o 


m o 


fN O 


o o 


o o 


o © 


d © 


o ©' 


3*8 












e 












O oe 












.2 «= 


fN fN 


>/-> oo 


SO O 


CN Os 


^r m 


E «« ** 


<— so 


OS CN 


os m 


TT fN 


Tf 00 


ompa 

of Mc 

P> 


Os — 


Os rj- 


rf rf 


00 OS 


fN fN 


</-> 00 


>r> SO 




SO O 


o r- 


o o 


o o 


o" © 


© O 


o o 


U 












h-Sill 
Rate 
iance 
ns) 












00 °. 


00 CN 


00 o 
© «n 


oo iri 


gs 


Witl 
Flow 
Var 


VO £ 


rr' * 


CN F"» 

(N — i 


>ri oo 


00 ~ 
























O B 












mm • « ••"v 


f*. V© 


oo oo 


29.77 

247.54 


8.79 
78.73 


17.18 
141.75 


e-Sill 

e Var 
(cms 


16.4 
55.9 


22.0 
16.0 


«- ■!; 












*3 












_ * « 












-Sil 
Flo 
cm: 


«•» & 


$£ 


CN t^ 


r*» r* 


Os </-> 


With 
Mean 
Rate( 


<o . 


Os 00 


sq Os 


fN -t 


<N 2 


CN 2 


© <ri 


© t*i 


— ' sb 


Sill 

Flow 

cms) 


00 °° 


gs 


oo 2 


<o m 


00 fN 


£ *» 08 


■<*- . 


r» os 


SO S 


rs JU 


CN 2 


M O 


©' so 


©' <o 


S o: 














ver 
iver 


ver 
iver 


ver 
iver 


ver 
iver 


ver 
iver 


c 


(5 « 


a* 


2 " 


2 K 


2 K 


OS 
05 


arys 
nnee 


arys 
nnee 


arys 
nnee 


arys 
nnee 


arys 
nnee 




2 % 


2 | 


2 | 


2 % 


2 | 




«J 3 


*J 3 


*J 3 


•J 3 


«J 3 




O0 oo 


oo O0 


00 00 


on oo 


00 00 






S? 




i- 


kn 


"3 

E 


I 

00 

3 


£ 
Q 

■*- 


u 

1 


E 

> 


E 
6 

o 




4> 


5 


o 


6 




< 


on 


o 


Z 


Q 



134 



in 

on 

On 

■ 

o 
<^> 

On 

c 



U 
oo 

c 

CO 

PL 

OO 

b 
U 

00 

en 

U 

E 

o 

> 

c 
o 

■ — 

.£• 



3 
o 

3 



c 



P 



c 
o 
9 
'C 

CM 

a. 

£ 
o 
U 

so 

1 



W 'I™ A 



c 
o « 

08 v A 

Ma. 

5 ° 



p* 


O 


r^ 


00 


— 


r^ 


O 


m 


ON 


NO 


r» 


on 


m 


•t 


r> 


SO 


NO 


ON 


ON 


* 


— 


^* 


«o 


^^ 


NO 


ir> 


m 


f-M 


VI 


in 


— 


"-- 1 


ON 


•—" 


CM 


m 


v> 


CO 


O 


— 


o 


o 


o 


o 


O 


o 


o 


o 


O 


o 



o ■<* oo o <n r^ 

(N io tj- r^ cn on 

On NO -—NO On (N 

On On T V> r^ «n 



o o o o 



o o 



(N CN 


o 


r- 


00 


<N 


rn r- 


00 


© 


>r> on 




o 




1/1 


© o 


r- 


© 


ON 


NO 


ON <n 


© 


00 


oo r- 


CM 


r- 


>o 




o o 


■3- 


oo 


■— 1 


m 


On On 


<N 


<N 


— »n 


NO 


00 


00 


ro 


© o 


r^ 


t-« 


m 


VN 


m m 


NO 


ON 


<N CM 


NO 


Tf 


«/-i 


•* 


o © 


o 


o 


o 


o 


o o 


© 


© 


© © 


© 


© 


© 


© 



i E2]>i 1 



_ >, .2 « 

— — +* W 

r3! ■* w» A C "— - 

f 8 3.S.5E 



L. 

a. 



-_ c 

f f 2 j « 

=T .5 08 *3 *•"< 

£E J « 

* - 



_ c 
k « o 

% I . J. 1 

ha j I 

* a. 



oo i— 






— < (N 



"3- — 



no © 



on <r> — r- v^ on 



•* © 

CN (N 



'tt© ONm oom —no ov >n t~-^ r^m 
CM(N (N (N m <n m<N rsfN — < — < rjrM 



<n cni 
on r^ 



r4 

On 

r<"i 



© ~* t> 
**■ no m 



NO (N 



■* © 

vO >n 



r^ Tt m Tt 



o oo 
iri no 



— oo 



NO no 



no so »r> no 



08 

w 



B 

o 





oo >n 




00 

© 


NO NO 

oo v-) 


oo 




— o 
on r- 


m r^ 


© 




m ^t 


f^i rf 


■* 


m 


m ^t 


r^i 


f 


■<* «n 


no r-- 


<o 


^D 





2 


a, 


S 


PL, 


2 


0- 
00 

b 


2 


(X 
00 


2 


a. 

00 


2 


b 
oo 


2 


a, 

00 

b 


2 


U 


U 


u u 


U C 


U C 


U U 


u u 


U u 


U U 


i/j 


C/J 


C/D 


C/J 


oo 


00 


oo 


oo 


00 


oo 


00 


00 


00 


00 


oo 


9i 



I 



'a 

< 



i 



«j 



3 



3 
< 



135 



"O 

a 

3 

c 
o 
o 

I 

so 

I 

CM 

o 

CO 

H 



rison 
ances 
F 


f- SO 


— i «N 


O 00 


© — 


On rn 


u-i S 


tt rn 


00 so 


S'u A 


r-- o 


tn r^ 


Tt r*- 


m oo 


o . 


— t^ 


Os t^ 


— o 


00 t^ 


o o 


o © 


o ©' 


o o 


3* 










c 










§ £ 

•pi H w 


V© _ 


wm -1 


so r- 


r-» iri 


U « 


</"> IT) 


Tt rm 


SO so 


Os >n 


M 4> A 

1*5 


<N — 


SO O 


m —< 


O </-> 


O O 


h» oo 


* -<t 


fN — 


o © 


© © 


© o 


© © 


U 










c 










S .2* « w 










With-S 
Biweek 

Total 
recipita 
Varian 

(cm) 


os m 


— © 


os n- 


V) — 


fN m 


rj- m 


t*\ rn 


<N (N 










a. 










c 










_ >> J « 










Pre-Sil 

Biweekl 

Total 

recipitat 

Varianc 

(cm) 


rf t"» 


O <N 


<r> o 


Tf Os 


■<*• m 


rf m 


CN <N 


Csl »»« 


>— . .— i 


•"* ^ 


— <— < 


— < — 


a. 










_ B 










— ■ **"' +-i •"• 


















ith-Si 
iweekl 
an To 

:ipitat 
(cm) 


m o 


r-» m 


CN f» 


m os 


Os «r> 
m' ^f 


o t- 
rn «n' 


— ; r^ 

en rn 


oo in 
rn Tt 


£ oa £ « 

* - 










_ c 










v « e 










Pre-Sil 

Biweekl 

Mean To 

Precipitat 

(cm) 


rsi m 


r^ cn 


Os O 


<n — 


Os r» 


»-| 00 

rn rn 


oo <o 
rsi rn 


rn tj-' 


c 










# o 






















oo 5 

Uh 02 


0* *f 


82 S 


oo S 

U. OS 




y u 


u u 


y u 


U U 




00 oo 


00 00 


oo oo 


00 00 


B 


1 


u> 


u. 

2 


lg 


o 


£ 


6 

X) 

o 


E 
B 


E 
6 




a. 


■*-» 


> 


o 




B 


Q 


o 


u 




00 


O 


Z 


Q 



136 
(Table 2-17). It is also likely that the increased river flow is due to greater flows in the 
creeks entering the Northwest swamp during this period, since creek flow would also be 
affected by changing evapotranspiration and precipitation rates in the watershed. 
Channelization and logging in the northwestern creek watersheds may also be affecting 
flow volumes and rates in the Suwannee River. 

Water levels at SCFSP and SCRA also changed following sill construction. 
Overall water depths were lower and more variable before the sill was built; this trend 
occurred during the growing and non-growing seasons, although the decreased variability 
in growing season water level at SCFSP was not significant following sill construction 
(Table 2-18). The smallest change in water depth occurred in October-November at 
SCFSP and September-December at SCRA, and variability decreased mainly during 
October-January at SCFSP and November-March at SCRA. At SCFSP water depths 
were higher during 10 months, and variability was lower during 4 months with the sill in 
place, but only December and January had both higher water levels and lower variability. 
At SCRA water levels were higher during 8 months, and variability was lower during 5 
months with the sill in place, but only January-March had both higher water levels and 
lower variability. These decreases in water level variabilities correspond to increased 
and more variable non-growing season (primarily January) precipitation recorded at 
SCFSP and SCRA while the sill was in operation (Table 2-16). 

Trends in biweekly rainfall totals exist among decades, although the differences 
among decades are not statistically significant (Table 2-19). Highest average biweekly 
precipitation totals recorded at SCRA and SCFSP occurred during the 1960s and 1970s, 



137 



I I. 

5-c a 

o r 

U o 



c 

£ »> 

E 05 ** 

S <_ 



= J c 

S 1 > « 






tf£S 

gE 



i S E 
life 

2 



O *C * MN* 

cN «n r<-i — < vo O 

o — vo o o o 

o o o o o o 

© O © O © O 



— vo (n 5 5 o 

in io in x s a 

cn «n o **-! H o 

© o d ^ ^ d 



r- m vo ■* <n o 

© »n © © © © 
d d d d d d 



ov ov vo m vo <n 

vo — oo — oo r» 

r- — w+ © ■* m 

cn o o o — — 



cn <n vo — oo oo 

m oo f» o (•»■ oo 

0O0O O O -(N 

«n VO O O O f> 



rsi r~ (N vo vo (N 

Tt- r«- m m - o 

(N a - Tt o m 

o »n o o o o 



oooooo oooooo oooooo 



O o 



N - M M 

o o o o 



oooooo 



o o o o § o 
d d ©' d - ©' 



m vo (n <N <n cn 

oooooo 

d d d d o* d 



>n oo <n 

o o o 

d d d d d © 



<n <n vo 

o © o 



CN (N — <N ~ <N 

OOOOOO 

© © ©' ©' o" © 



vo t"» <n in vo t** 

OOOOOO 

©do© d d 



ov vo t m >n o 
d d o* d d o" 



— VO 00 O t"» <N 

r^ oc i(s oo r- t^ 

Ov Ov Ov Ov Ov Ov 



00 0O rf <n Ov * 
f*; O f^ t*- VO 00 



1 -^ 

ESS? 

? bl 

ft. c H 

o 

2 



B 

C 

« 



c 
o 



S^^Srs- £ t- vo cn cn <n 

*_! *J °1 °1 . f^ 00 00 Ov 0C f- 

2 2 2 2 2 2 ov' ov' ov' ov' os o< 



o >n vo «n >n vo 

ON O 00 00 00 Ov 




c 

O 

_ *-» 
O co 

oo.* 



u. 



1 

E 

o 



o 



3 
c 
■s 

<a 
Z 



C/3 




;__: 


t/5 

£ o 


a 



Homervi 
Nahunta 


5 5: 


5/5 
k> — 


££ 


CO O 



w 




z 




•* 




C/l 




V) 




£ 


O 





2 


a 


C/) 


££ 



CO 

3 
C 
C3 



ca 

2 

X) 

(L> 

— 



1 



138 



T3 

1> 
3 
O 

c 
o 
u 



(N 

x> 
H 



I 8 

■S e 


























b. 


o 


o 
o 
o 


o o 
o o 
o o 


11 


rn 


00 


on in 


ON ON 


■* 


t"* 


O O ON fN 


§ « 


o 


o 


F» no 


o rj 


r^ 


On 


"* — O — 


« -p. 

- 5 

I> 

e 


A 


o 


o 


r^ 


»-* »— i 


o o 


(N 


O 


— < O — < (N 


a. 






o 


o 


NO O 


o o 


O 


1—1 


r~ 00 




© 
V 


o 
V 


o © 
V V 


°? 


o 


© 


© © 


o o 


O 


O 


0000 


e 

S ■ 
•- = 




























— 


p-I 


On oo 


On — 


m 


© 


v~t m 


<^> o 


_ 


ON 


Tt 00 — - ~H 


C ee 


** 


o 


© 


<T) fN 


O O 


ON 


^1 


en o 


00 On 


O 


O 


tt 00 


I* 

e e 


A 


o 


o 


»— * »— I 


o o 


NO 


en 


NO NO 


r» F- 


O 


O 


0000 


0. 


o 


o 


o o 


o © 


NO 


o 


O On 


(N '3- 


O 


O 


0000 




© 


© 


© © 


© © 


© 


© 


O O 


o o 


O 


O 


0000 


U 


























§ £ 


























_ ■ " 


























h-Sil 
lyM 
ariar 


** 


o 


as 


o fN 


fN NO 


On 


00 


^r 00 


o — 


■>* 


en 


(N r^ m no 


£ 


V) 


© 


cn m 


fN «r> 


m 


o 


«-« ^-< 


m m 


Cn| 


>-* 


^- f^ (N (N 




U 


© 


© 


o © 


o © 


© 


© 


o" o 


o o 








O O O O 


























? 


























Pre-Sill 
Dnthly Mean 
T Variance 


? 


NO 


r- 


no t-» 


</-> no 


o 


w-i 


P* 00 


00 o 





m 


nto\« 


o 


V) 


o o 


o o 


•— 1 


— 


— o 


O — I 




CM 


-MO- 


o 


o 


o 


o © 


o © 


© 


© 


o o 


o o 





O 





S u 


























■ 


























I 


























— »> ^ 

81 s E 




























TT 


On 


— Tj- 


On — 


00 


ro 


«n m 


NO * 


<n 


>r> 


« r> 


J.*l 




00 


(N 


© On 


0s oo 


00 


rr 


tj- m 


^t o 





(N 


On r~; r>- (N 


lib 




m' 


m' 


rt m' 


en" rn 


"♦ 


iri 


TT TJ-' 


rf vS 


NO 


NO 


</i <ri v~> no 


plfi 


























s 


























s 


























■ i? 




■*r 


NO 


r-» (n 


■x NO 


ri 


"* 


* m 


00 NO 


«r> 


(N 


>n r- — 








fN 


fN fN 


m no 


oo 


fN 


NO W-l 
TT Tt 


>n On 


NO 

NO 


NO 

SO 


NO NM 

NO NO NO NO 


ft. s u 

o 


























5 


















































s 












z 








2 






2 


e 
e 
•a 








J£ 


Ti- 
en 






Ji 


n 






9 to 


I 
in 






C 

o 


li 


■B 

20 




C 

o 


ll 


s o 




c 



Ills 

ac z ^ ^ 






O v} 

eo o 


11 

O CO 




o tfj 
oo.* 

t- — . 
<fl o 

Ll. U- 


il 

O RJ 

a z 


5 5: 


O 'K 

£? — 

C3 O 


e 
e 

z 




c 
B. 

< 








& 

2 








c 

i— » 







139 



•o 

3 

5 

B 
O 






1 2 






















1 u 

•- S 


h 


© 


r^ 


P o 


o o 
o o 
o o 


_ 


Ov 


(N ^h — < O 


<#> p-< 


vo * m "t 


£ « 


o 
o 


00 


is 


m 


F» 


»— >n m r\i 


•O Tj- 


otoa> 


re -p- 
it 

U o 


A 


r- 


o 


r- 


rN m o — 


m — < 


ov <n oo rs 


a. 


o 


o 


i-^ 


Tf rr o — 


o o 


</-> t*> o o 




o 

V 


o 


<=>•? 


d o 
V V 


o 


© 


d d d d 


d d 


d d d d 


B 






















s ■ 

•- = 




'* 


© 
o 
© 


5 Tf 


«N «• 


«n 


Ov 


rf a (S t 


vo >n 


© oo 5f rs 


- re 


** 


(N 


■f 00 


•— *n 


00 


o 


o *fr (N m 


vO O 


- Tf o\n 


re v 

ft 

e e 


A 


r» 


O r}- 


© o 


o 


VO 


o — S ■* 


<N O 


1^- 1- VO (N 


C 


o 


F tj- 


Ov O 


VO 


— 


o o o oo 


VO O 


*■ O O Ov 




© 


o 
v 


o © 


© © 


© 


© 


o d d d 


d d 


d <z> d c> 


U 






















§ £ 






















- o i 






















h-Sil 
lyM 
ariar 


-. 


T 


On 


rn m 


Tf 00 


(N 


in 


OOOMN 


On l^ 


rt m in »- 


E 


■* 


O 


rn t 


<n «n 


•■< 


o 


— O — < — 


r%i o 


— — cn m 


WU 
9 nth 
TV 


B 


© 


o 


© © 


o o 


o 


© 


O d d d 


O d 


O d © d 






















s w 






















I a 






















Pre-Sill 
onthly Me 
T Varianc 





vo 


r^ 


<N VO 


>n VO 


VO 


Ov 


oo vo vo r- 


ro vo 


rN — rs tj- 


E 


© 


^ 


— o 


o o 


O 


o 


o o o o 






' — ' 


© 


© 


© o 


© © 


© 


d 


<z> d d <z> 


d o 


d d d d 


S w 






















■ 






















St 






















i^ e 
























Os 


m 


*— <n 


Tf <N 


o 


00 


rr t*~ t> oo 


>n oo 


in —> vo — 


X i>w 




CN 


"~" 


r- t^- 


p«. vq 


00 


*-" 


(N (N ft OS 


rn o 


— ; c i p «n 


1 £ H 




vd 


r->' 


»n «n 


«n vd 


<n 


vd 


m' «ri iri in' 


rn tt 


rn m m rn 


£ B« 






















z 






















E 






















■ 






















IS E 




vO 


■>* 


<n »n 


m rs 


"* 


Ov 


rM tj- rN ov 


© vo 


rs os >n © 


Pre-S 
nthly 
ET(c 




o 
vd 


vd 


>n <n 


S <N 

<n vd 


00 

<n 


o 
vd 


VI -rt <r> Ov 

«ri iri «ri in' 


5F r-- 


rs — > (N m 
rn rn tn rn 


e 






















£ 
































g 






g 




g 


B 
■ 








jd 








13 co 




— CO 








c 
o 


1| 

il 

O C3 


CO 

s 9 




c 
o 


Ills 


c 
o 


Ills 

11 IS 






O co 

SX) -^ 

eS o 




O To 
CO o 


il £2 
S 3 5 g 

I z ^ ^ 


O co 
Ofl^! 
u. — 
CO O 

6 

JO 






















B 
■ 












CO 






E 




2 




3 








I 

3 

< 






u 

■♦-* 
a 
<u 

00 





140 



3 



C 
O 



CN 

H 



1 £ 

- 33 




CN 


o 
o 
o 


a: 


«r> co 


CN 


ON 


o 




>/-> 


CN 


CN 


•<i- 


— 

o 
o 
o 


t-» 


CN 


00 NO 


hi 


«n 


o 


— Tf 


m 


ON 


^— 


* 


u-i 


«o 


O 


0m 


*r\ 


r*- 


© f- 


SL'C 


A 


o 


* 


00 o 


o 


O 


00 


r» 


^-1 


S 


CN 


i— « 


co 


r^ 


ro © 


=■ « 


S. 


o 




CN •— 


o 


CN 


cn 


On 


"«■ 


o 


On 


o 


V) 


o 


© © 




© 


© 
V 


© 


© O 


© 


© 


o 


© 


O 


o 


d 


o 


d 
V 


© 


© 


d ©' 


e 




































.£ 1 




pM 


M 


© 

o 
o 


II 


__ 


p« 


rf 


tT 


_ 


r^ 


CO 


pM 


pa 


■<i- 


CN 


5* — 


C re 


-— 


o 


o 


o 


r-» 


^^ 


«* 


*— « 


On 


CN 


o 


O 


© 


O 


o © 


re 4< 


A 


o 


o 


o 


00 


CM 


o 


ON 


vm 


On 


o 


O 


© 


© 


© © 


e«S 


C 


o 


o 


o 


*; 


co 


CN 


CN 


r^ 


On 


o 


O 


© 


© 


o © 


E «_ 
e o 




© 


o 


© 
v 


?<=>' 


© 


o 


© 


© 


O 


o 


d 


d 


d 


© 


© 


© © 


U 




































re c 

-Si 




































h-Sil 
lyM 

a riai 


w 


^* 


On 


co 


On CN 


rf 


co 


NO 


TT 


m 


ON 


NO 


o 


CN 


© 


rr 


*T) CN 


E 


<N 


o 


•— i 


O — 


cn 


o 


o 


o 


o 


o 


o 


CN 


O 


•— < 


"— 1 


— < CN 


sf > 




o 


© 


© 


© © 


© 


© 


© 


© 


o 


o 


d 


d 


© 


© 


© 


© © 


* §H 




































s w 




































Is 




































re-Sill 
thly M< 
Varian 


^ 


r- 


SO 


r~ 


NO NO 


00 


<r> 


»o 


Tr 


TT 


* 


«o 


00 


r- 


00 


f» 


r^ oo 


s 


o 


■<3- 


o 


o o 


o 


o 


o 


o 


o 


o 


o 


o 


CO 


© 


© 


© © 


o 


© 


© 


© 


© o 


© 


o 


© 


© 


© 


o 


d 


© 


© 


© 


© 


© © 


lb 




































s M 




































e 

09 




































— ] ^ 

S 2 £ 






































© 


co 


'S- 


00 o 


*— 


NO 


o 


(N 


00 


•* 


CO 


Tf 


On 
On 


NO 


f 


NO © 


.r -= i_ 




CN 


CN 


CN 


<n no 
cn cn 


cn 


© 


o 


CN 

o 


CN 

o 


CN 

o 


CN 

d 


co 

o 


© 


© 


•^ CO 

© © 


1 I" 




































s 




































e 




































■ 

4* 




































£ S I 




w-> 


<ri 


H 


m ro 


t 


o 


>r> 


On 


CN 


r^ 


ro 


CO 


On 


Tt 


r^ 


NO 00 


5 »> 5. 

£2 H 




© 
co 


00 
CN 


On 
CN 


On On 
CN CN 


rn 


CN 
O 


co 

© 


CN 

O 


m 

o 


CN 

d 


CN 

d 


00 

d 


© 


00 

© 


00 

© 


oo oo 
© © 


fi IE 




































e 




































2 














































1 












g 












w 
z 


e 
e 








U 


^f 








u 




f 








a 




■5T 




































a 










00 












(A 












V) 








C 

o 


| 


a 1 


O 




c 
o 


1 


i 


s 


O 




a 
o 


1 


a 

c 

3 
JC 

Z 


s o 






o x 

00.* 

3 "o 


1 J &? 

5 "S 3 ^ 

as z £ £ 


O T75 
00 jSj 

03 O 


1> — 
E | 

O C3 

PC Z 


5 ^ 


O Tfl 

I- — 

(3 O 

— Ll. 


o 

E 
o 


S 5 


£ 




u 










G 

25 












E 

,2 










1 
e 












E 












E 










s 




o 

O 










> 
o 

Z 












6 

u 
u 

Q 











141 



C 


<n 


e3 


os 


O* 


ON 


OO 


1 


L_ 


^— 


u 

oo 




<*. 




o 


_~ 






c 


M 


o 
C 


l- 

5 


C3 




a 


a: 


E 


o 


o 


u 


U 


g 




9 


00 


5 

n 


CN 


OO 


« 


M 


£> 


M 

•—> 


S3 


'-*- 


H 


o 



1 8 








o 
o 
o 


rr 


r^ 


o 
o 
o 


o 
o 
o 


— CN 


r~ 


r-» 


o 


rs 


00 


r-» 


SO O 


li 




O 


<r> 


o 


© CN 


SO 


SO 


r- 


^H 


00 


so 


</-> so 


«t 


A 




O 


o 


o 


o o 


m 


o 


o 


o 


«o 


(N 


— rs 


e * 






o 


*-^ 


o 


o o 


«r^ 


o 


*-^ 


o 


«-H 


^1 


00 — < 


on 


— 




© 


© 
V 


d 


d 


d 

V 


d 

V 


d d 


d 


d 


d 


d 


d 


d 


d d 


■ 


































O K 










^— 










_ 




^_ 




„_ 




_M 


.2 e 






^^ 


^-" 


o 
o 
o 


^H 


^H 


SO 


CN O 


o 
o 
o 


'* 


o 
o 

o 


IT) 


o 

o 
o 


so 


§8 


C es 


"** 




o 


o 


O 


O 


Tj" 


O >/"> 


»— * 


^H 


<N 


s 4* 


A 




o 


o 


o 


o 


— 


O 00 


o 


o 


o 


a. 




o 


o 


o 


o 


O 


o o 


o 


o 


o 


e © 

c 






o 


o 


© 

V 


d 


d 


d 


d d 


d 
V 


d 


d 
V 


d 


d 
V 


d 


?o 


Pre-Sill 
onthly Mean 
ater Surface 


c 
1 
•c 




so 


TT 


so 


TT 


SO 


rt 


tt co 


■<* 


m 


tt 


<N 


so 


co 


m m 


> 

B 


I 

•c 

1 


o 
© 


o 
o 


© 
© 


o 
d 


o 
d 


o 
d 


o o 
d d 


o 

d 


o 
d 


o 
d 


o 
d 


o 
d 


o 
d 


o o 
d d 


S£ 




> 






























Pre-Sill 
onthly Mean 
ater Surface 


c 
o 


1 






























s 


b 


00 


r- 


oo 


SO 


o 


r- 


•m r^ 


«r> 


so 


t^ 


SO 


"3- 


in 


m •'T 


I 


c 


o 


o 


o 


O 


»-" 


o 


— o 


o 


o 


o 


O 


O 


o 


o o 


> 

H 


re 
re 


© 


© 


d 


d 


d 


d 


d d 


d 


d 


d 


d 


d 


d 


d d 


S£ 




> 






























i? i 


































.e o 


































Monti 
Water 
Elevati 


C/3 




o 


so 


o 


so 


Os 


*■ 


co so 


<n 


m 


Os 


so 


(N 


rn 


so m 


2 




o 


v> 


o 


«r» 


OS 


>/-> 


o v> 


•— < 


SO 


<-" 


SO 


— •* 


SO 


OS «o 


— = 4. 


< 




vi 


SO 


</■{ 


sb 


■<* 


so 


<n sb 


«/S 


sb 


<ri 


sb 


«ri 


SO 


Tf sb 


t« ? s 






ro 


co 


CO 


co 


co 


co 


co <^> 


f«i 


r^ 


m 


m 


m 


m 


m ro 


■ « re 


c 
































£2t 


































£ 3 

> CA. 


































* § 


































5 B a 


J 
































£ £ £ 


n 




Os 


00 


o 


00 


Os 


00 


OS 00 


(N 


Os 


r-- 


(N 


«o 


»— 


rr t3- 


£ £ S 


£ 




r- 


TT 


00 


tt 


r«- 


"♦ 


F« ■* 


00 


^r 


00 


«ri 


00 


«r> 


r- •>* 


■ I ■ 


< 




rr 


SO 


Tf' 


SO 


t' 


SO 


rf so 


•<*■' 


SO 


rr' 


sb 


rf 


sb 


rt b 


ass 


E 




CO 


CO 


to 


CO 


co 


co 


m m 


m 


m 


m 


w 


m 


ro 


ro m 


Mi 


































£ & 


































■ 

e 


































••c 

■ 

8 






0- 

00 


2 


Oh 
00 


2 


a. 

00 


2 


00 *S 


a- 
oo 

Pu 


2 


a- 

oo 


2 


OO 


2 










u u 


u 


U 


u 


U U 


U U 


U U 


U U 


U U 








00 


00 


00 


00 


oo 


oo 


OO 00 


00 


00 


oo 


00 


00 


oo 


oo oo 


"5 

■ 
1 


1 

D 

(5 




00 

c 

o 
c 

o 


a 
C 

o 

C/5 

03 
<U 

oo 


1 

c 

O 


OX) 

c e 
'5 9 

Eboo 


3 
C 

*- > 


1 






CO 

2 




< 




2 



142 



3 
9 

c 
o 
o 

t 
I 

00 



CS 
Jg 

OS 

H 



rison 
ances 
F 
















SO >-< 


m v> 


CN On 


rf u-> 


On cn 


— CS 


^- r-- 


O O 


<n m 


— CS 


CS O 


cn t** 


o cn 


•<* o 


Vari 
P> 


O 00 


— Wl 


00 NO 


v-> cn 


— cn 


o cn 


o <— 


r- cs 


V^ On 


_l fsj 


cn m 


o — 


o o 


o o 


© d 


© o 


© © 


© © 


d d 


d d 


d d 


W "S 
















■ 
















e cn 




Si 


— O 


00 *T) 


00 On 


ON TT 


<n c^ 


— o 


t> 00 


oo cn 


— i vo 


cs K 


§■5 a. 
5 e 


O NO 


— ON 


O 00 


00 CS 


— cs 


• o 


o © 


o m 


r^ r^ 


— K 


o m 


?p 


?<=>' 


o o 


© © 


d d 


d d 


d d 


B ■ _ 
















Mil 

Mea 
urfac 
tion 

e (m 
















VI Tt 


>n *j- 


k> >/-> 


00 NO 


r- no 


»n <o 


NO ■** 


Pre-! 

aterS 
Eleva 

ariam 


o o 


o o 


o o 


o o 


o o 


o o 


o o 


© o 


o o 


© o 


d © 


d d 


d d 


d d 


S£ > 
















B * ,— 
















- 1 1 1 1 
















<r> v> 


NO rr 


o r-- 


O 00 


cn On 


^t ON 


CS On 


if til 


o o 


o o 


T- o 


— o 


— o 


— o 


— O 


© o 


© o 


o © 


d © 


d d 


d o 


d d 


e R u « 
















S* > 
















X> 1 
















.= c 
















t: -_ •- 
















1 Mon 
Wate 
Elevai 

MSL) 


— ON 


5C cs 


On r-~ 


ON 00 


cn cn 


r- oo 


CS ON 


On tJ- 


On ■/"! 


On *r> 


On </-> 


On V) 


00 rj- 


ON -3- 


£ S K < 


rf NO 


■^■' vd 


■<*•' NO 


■» NO 


rr' SO 


Tf' NO 


^r no 


? 2 | E 

ill* 


m cn 


m m 


m m 


cn m 


cn cn 


cn cn 


c^> cn 
















? Cfl 
















* § 
















JS - -J= 
















Mont 
Wate 
Elevai 

MSL) 
















r- r- 


jt ^r 


— « On 


«/-> m 


<o ^r 


00 On 


NO NO 


no m 


k> rr 


00 •«3- 


oo «n 


00 v-i 


l> TT 


t-> rt- 


re-Sill 

Mean 

irface 

(mA 


Tf' NO 


rr' no" 


Tf' NO 


Tf' sd 


TJ-" NO 


TJ-' NO 


rt NO 


CO o 


cn cn 


m m 


cn cn 


cn cn 


cn cn 


cn cn 
















*■ Cfl 
















B 
















e 
















SB 
■ 

en 


01 *f 
oo "s 
tt, ad 


0i -f 

00 "S 
b 05 


0i «*- 
OO "s 


0- <f 


0i *f 


0- < 


0i *f 

00 <, 
(X, K 




u u 


D u 


U U 


y u 


U U 


u u 


u u 




OO OO 


OO OO 


oo oo 


00 00 


oo oo 


oo oo 


00 00 


1 
fc 

£ 








1 

s 




h 


1 


u. 
X) 






to 


| 


E 


E 


B 


a 


"3 


3 


o 
B 


1 


B 




>—> 


<-» 


< 


oo 


o 


2 


Q 



0) 

E 

D 
O 

u 

Q 
i 

i- 

X) 

£ 

> 

o 

c 
5 

X) 

u- 

I 

rt 

3 
C 

<a 

<—> 

cn 

<U 

T3 

^3 

O 

c 

c 
o 

C/5 

rt 

<u 

00 

c 



o 

I 

c 
o 

c 



X) 

o 
o 

o 

CO 

<u 

"O 

^3 

o 

_c 

c 
o 

cn 
rt 

(U 

</i 
(so 

C 



O 

























o 


















143 




« 




On 




On 




r^- 




00 


z 


on 

ON 




© 




© 




p 




p 






^ 




*»* 




t-i 




T-* 




v-^ 




M 

■a 


l 


O 


1 

NO 


© 


1 

NO 




■4 


<N 


■4 




I 
























i 


© 


1 


© 


■ 


d 


i 


d 




£ 








































.1 


^ 




(N 




On 




w 




w- 




1 


00 
©\ 

i 

o> 




*— < 








•"! 




*~. 




J 
1 


© 


00 


© 


i 

oo 

00 


NO 

O 


00 


NO 

O 


i 

1-H 

00 




"* 


1 


o" 




© 


i 


d 


*7 


d 




o 






















o 
























On 




















in 


t~ 




On 




r*- 




00 








On aj 


On 








fS 




<— < 






■a a 

C TO 

5 E 


i 

e 
r~ 

On 


(N 


1-H 


00 


1 


o 
o 


1 












O 


00 


© 


On 




00 


. 






CRA 
were 




— ' 


© 


— 


© 


1 


d 


i 
i 


1 






















00 B 






















C 


On 




















"2 o 


v© 




On 




r^ 










CO T" 


On 








<N 












C/3 E" 


i 

On 


(N 


i 


00 


i 












S E 




© 


00 


© 


On 


1 


i 

1 








Q R 




w 


© 


1— 1 


© 


1 

1 








00 u 










































TO g 

en Or 


On 




















o u 


If) 




o 
















■o JO 


On 




1—1 
















"O .y 


i 

On 




«-H 
00 




i 












Mia 
C C 




© 


© 


1 
1 












o F 






















E o 






















TO C 


On 




















C 00 


On 




















o o 
























i 




















TO 0> 


© 






















On 




















■^ 


1 
1 


i 
















U — 




1 


1 
















6 5 








































D. 5 






















>> Q 




V) 




tfl 




(A 




00 






"5 «"i 


L 


c 




§ 




e3 




§ 






c o 


■ 


D 




D 









8 






C VI 


1 

E 

1 






0> S 




i c 




8! 

c c 






03 o 


I 
C 




»-H 

U 




u 


K 1 


— 

U 




U 




in me 
;ant at 


Hi 

Q ffl s£ 


IT) 

On 


HI 


s - 

On 


If a 


s - 

On 


ill 


s - 

On 




Vi *-» 






















U <s 


























On 




On 




On 




ON 






c '5 

O QQ 

~ — 


1 


On 




£ 




« 




On 






f i t/i 


^"< 




~- 




^H 




»-• 






Diff. 
were 


** 

s 


1 

o 

5 




© 




S 




O 
P. 








o 




On 




On 




ON 








^H 




^H 




i-H 




»— H 






2-19. 
ences 






















■ 

e 


Oh 
00 
















able 
iffer 


"•O 

■ 
09 


U 
00 


















M -O 































































-o 

<D 

3 
C 

5 
o 
o 

I 
I 

as 



X) 

CO 

H 































On 

ON 






m 






m 














<* 






*-; 






p 








•— | 






p 


i 






«— < 






*— i 








— I 






M 


s 


NO 

o 




00 


On 




1 


•* 






On 


m 




t 


ON 






1"» 


oo 




t- 


On 






t»~ 


V— 4 




r» 
































1 




© 


O 




© 


© 






o 


•"' 




© 


On 
00 












»— i 








r- 






o 


On 












^^ 








*-* 






»— « 


** 

s 












— i 








— ! 






KM 








VO 




4 


(N 






On 


v> 




(N 


ON 


1 
1 
1 




i 
i 
i 


On 

o" 




oo 
© 


O 






00 
© 


On 

O 




oo 
© 


on 




























t^ 












oo 








m 






r^ 


On 












«~J 








(N 






*— 


© 












»— < 








~ 






PM 










<N 




1 

00 


00 






i 

m 


O 




1 










© 




oo 
o 


© 






On 
© 


O 




00 

©' 


on 




























9 












00 








VN 








On 












~« 








(N 








On 












•-! 








-■ i 














(N 




00 


00 






m 
















© 




00 

© 


p 






ON 

o 


i 




1 
1 
1 


On 




























V) 

ON 












8 
















l 












1—1 
1 
















On 








«/-> 




(N 
























On 
© 




00 

O 


: 

1 






■ 
■ 
■ 








On 




























^ 




























On 
























































e 




























•«• 




























On 




































i 




1 


















1 






8 








| 






g 






k 


CO 






a 








cd 






Ctt 






1 


D 






S 








<u 






fll 






S 


R 2 






o>2 






u 

o 


2 












2 


c c 






c P 






c 


c 






c c 








i- u 




u 


0) <U 




U 


6 






u 


5 <u 




— 

u 


I i 




o N 
On 


II 

Q CO 


? 


On 


ij 

Q CQ 


? 




m 

On 


il 

Q CQ 




On 




a 






On 






On 








On 






1 
t 


00 
On 






a 






On 








NO 

On 






•■H 






•— 






— 














■ 


1 






1 






1 








| 






B 


o 






O 






O 








o 








00 
On 






I 

1— 1 






On 








NO 

On 






■ 




























O 








^f 




















-c 








-u) 




















03 








— 




















s 




























■ 








35 





















144 



145 






a 
o 



ON 

rs 

M 

x> 

H 







s 




On 
O 


i 




— t 




-h' 


? 


t; 


m 


CN 


pl- 


C 1 


00 


K 


ON 


r^ 














© 


© 


© 


o 


©n 
ae 
on 




o 






^* 




^H 






© 




1 






ae 


>o 


CNI 






m 


On 


00 






~ 


© 


o 


■ 




©\ 










t~- 










« 










pn 










© 










F 










a 










«* 


i 
i 
i 


! 

1 






ON 










i 










^* 










© 










! 




















A 










$ 










IN 










2 










v> 










ON 




















©v 










<* 










ON 










«N 










6 










■* 










ON 






















1 




g 




h 


ea 




rt 




B 


y 




<o 




E 


« 2 




N 2 




■ 
a. 


c c 

Si 5J 


U 


c B 
Si u 







s - 

On 


lis 

Q ffl 3 


2fi 

On 




Os 




ON 




-a 
E 


F» 




00 




On 




ON 




•^ 




«— < 




1 


© 




© 




F 




oo 






On 




ON 






•"" ' 




*"* 




c 










e 










•« 










S 




















55 











146 
and lowest totals occurred during the 1930s (possibly due to continent- wide drought 
conditions of the "dust bowl" era) and 1980s. Biweekly precipitation totals were 
intermediate during the 1940s, 1950s, and 1990s. At SCFSP water levels were highest 
during the 1960s and 1970s, lowest during the 1940s, 1950s, and 1980s, and intermediate 
during the 1990s (Table 2-20). At SCRA water levels were highest during the 1970s and 
1990s, lowest during the 1950s and 1980s, and intermediate during the 1940s and 1960s. 
During the with-sill period, high and medium water level periods have corresponded to 
periods of high precipitation, whereas lower water levels occurred prior to the sill's 
construction when average precipitation volumes were also lower. The sill's affect 
appears to be mainly during high water and high precipitation periods, and when 
precipitation decreases, water levels in the sill area (represented by SCFSP) and 
throughout the swamp (represented by SCRA) also decrease (Figure 2-17). This 
indicates that the intended purpose of the Suwannee River Sill "to prevent drainage of 
the Okefenokee Swamp during periods of drought" may not be achievable with the 
existing sill configuration, and correlation of swamp water level and precipitation 
volume. 
Topography Surface 

The swamp topographic surface was interpolated from elevation data collected by 
4 methods: Global Positioning System (GPS) survey (106 points), laser transit survey (48 
points), "flatpool" survey (498 points), and USGS 7.5" 1:25,000 topographic quadrangles 
(362 points). Elevations above mean sea level (AMSL) representing the peat and 









































































147 


g 


— 








I/I 




© 




© 




© 


1 




O 




© 




d 




l" 




© 
1 


y 


s 




' 




1 




1 


* 


1 




•3 

i 

irs 


- 


o 

© 


© 
o 

1 


# 

d 


On 
O 


O 

d 
i 


© 
© 

1 


© 
i 


00 

d 

1 


© 

d 


SO 

9 


ON 

ae 












s 




o 
o 






On 


■ 




© 




d 




© 




1 






c 


OS 


o 


o 


• 


1 

ON 


o 


— 


• 


Os 






£ 




© 


© 

1 


d 


d 


d 
i 


© 

1 


©' 


o 

1 


1 


I 


~ 






















00 


r~ 






















"2 


os 

1—4 




s 




9 




wi 










§ 


1 

o 




© 




d 




©' 










t- 


* 


1 


* 


i 


• 


1 










Oh 


©s 


■> 


1—1 


r- 


tm 


00 


p4 










C/D 




*— 


— 


i*> 


r"> 


© 


o 


1 


1 






I 




© 


d 


© 


d 


d 


©' 


1 
1 


1 

1 




























3 


ON 




no 




SO 














PN 




"■" 




M 














"2 


© 




d 




d 












§ 


NO 

Os 


• 

OS 


irs 


* 
o 


i 

c*s 














•s 




© 


o 


tr> 


(S 


■ 


t 

1 












© 


o 


© 


d 




1 

1 










V 














































o 


Os 




fs 


















c 


P4 




© 


















o 


© 
























v. 


* 




















evat 
0.05 


ON 


o 


r~ 


















i-» 


© 

1 


© 
• 


1 
1 

1 


! 

1 














« VI 






































































:r surfa 
cant at 


os 
©s 




















1 






















a is 


©\ 




















i a 




1 
1 
1 


i 
i 


















>% 55 






































































32 <L> 
























0) W 
























£ * 




1 




i 




I 




i 




1 




in mean 
arked wil 


E 

2 

a 

0. 


1 § 


U 


8 2 
l! 


U 


a> 5 
If 


— 


8 if 


u 


8 s 

B H 


D 


1 € E 
Q CQ S 


ID 

On 


ili 

S 05 S 


•s? 

<ys 

On 


IS? 


>5 

Os 


£ 13 E 
Q CQ 5. 


in 

On 


IS 1 E 

a cq ^ 


Os 


4) c 














































O W 
























C O 
























Si o 
























20. Differ 
. Differen 


iS 


Os 




OS 




ON 




On 




Os 




i 


On 




Os 




NO 

Os 






00 

Os 




e 


i 

O 
rr 

On 




6 

Os 




1 

© 

NO 

Os 




i 




i 

© 
00 

OS 






»— ■ 




■""■ 




*"* 




^^ 




•■* 


























1 V} 
























<N « 


e 
o 


a. 




















•s i 


- 


t/5 

E 




















X> <u 


~ 


o 




















1 s 
h .a 


wj 


«/3 













































148 



3 
C 

i 

o 



© 

i 
1 



f»> 






















o> 

on 




§ 




s 




§ 




s 




<N 


1 




© 






© 




©' 




©' 


1— < 


(N 

o 


© 


# 
V) 


i 

oo 
© 


© 


© 


el 
© 


© 


© 


© 




©' 


o 

■ 


©' 


© 


©' 


© 


©' 
i 


©' 

■ 


© 


© 


©N 
00 
On 




(N 
O 
©' 




v-, 
©' 




© 
© 




© 
9 






© 




i 




i 




■ 


• 


1 






ae 


5» 


o> 


* 




u"> 


© 


00 


, ^- 






©V 


© 


© 


On 

o 


§ 


© 




© 


»-* 






^ 


©' 

t 


©' 
i 


O 


© 


O 


© 
• 


©' 

1 


© 

1 


i 
i 


i 
i 


On 






















On 




~- 




3 




S 










1 

© 




© 




©' 




© 










l-~ 








1 




' 










On 


o 


o 


00 


© 


© 


© 












©' 


o 

1 


© 


©' 


© 


O 
i 




1 






OS 






















VO 
On 




£ 




© 






















<s 














i 




© 




©' 














On 


© 
©' 


O 

o 
i 


* 

©' 


oo 
© 
©' 


i 
i 


i 
i 

i 










ON 
V, 

On 




S 






















© 


















© 




' 


















V) 


• 


' 


















o. 


rn 


On 




















©' 

1 


O 
i 


1 
1 
1 


i 














ON 






















■t 






















ON 












































© 






















If 






















ON 
























1 

I 
1 


i 
i 
i 


















1 


! 




8 




i 




1 




J 




E 

2 

« 


B - 

s § 

ill 


— 


8 * 
1 1 


<— 

U 


8 s 
s ! 


u 


V ft 


u 


U 




>5 


£ It 

■ _ -j c 




1 £ e 


^5 


1 I £ 


6 s 


*£ £ e 

Q S3 £, 


4 

o^ 




Q 03 ^ 


ON 


a pq S. 


ON 


Q«^ 


VN 

On 


Q OQ 4i 


NO 

ON 


NT) 
ON 


























On 

a; 




ON 




ON 
NO 

ON 




1 




ON 

00 
ON 




e 


© 

ON 




© 




© 

2 




i 




© 

00 

ON 




e 






















o 
- 


2 




















t/3 























149 



SdOS }v uo!;ba9|3 aoejjns J9JBAA ^mi uo lAI aBejaAV 



•«■ 


CM 00 


CO Tt 




W 


W IO ^- 


•* ^ 




CO 


CO CO CO 


CO c 




o 






! 1 1 


1 






^ 




0) 










0) 






" r:::::: --v. '^r 








/ — ^- ■■::::::::::::: ' 






c 




—:-.::;;: " --~^. 




5 


o 




_-' ^^"\ 




^ 


53 
CO 












-- ~~3*-..- 




o 

00 


> 








O) 


LU 








^ 


CD 
O 








m 


i? 








0) 


3 
CO 


<^~"~ v. w 










u 




, — ^^ V:::::. 




o 


0) 




£ , __ . t ; 




h- 


*j 




-"^^V^." 




O) Q. 


CQ 




~~^^- "\ 




O 


£ 








c 


i 




^y ..'" 




IT) rt\ 


I 


"""■"~ 




— — _•-"-" :::^p ::: """ 




60 19 
5-Year 




^m- "-*' 






^- "" "•'-'-'•"."." 




O) 






™ — ~HX "-'■' 




*" 






"'''■'-:•..- / 




in 






-''• ' """ J 




IT) 


c 




• : -.^ C- 




0) 


o 




^-.-/." 




T- 


53 




•■===:::::::;;;^: "*"< 






s 




....1— — — ~~" ~— 




o 
in 


a 








0> 


o 








T- 


a) 










0. 




. ™ T-^ 




m 










•t 






■ ** 




o> 






^—^ ^ 4 




o 










1 1 1 


1 


5P 





00 



CM 



(luo) dSdOS »e uone^idpdJd A|q}uo|/\i aBejaAV 



§5 
& 

1 






a 

§ 

1 



1 

I 

■ — 

I 



c 




o 




£ 




0> 




au 




g 




c> 


* 


> 


On 


< 


ON 


s 


"• 


1 


On 


<N 


*-* 


4> 


an 


t- 


c 


ap 


3 



Ph *0 






150 
underlying sand surfaces, and the thickness of the peat on the sand surface, were 
calculated and included in the data sets, which were interpolated to create data grids. 
The sand and peat surface grids were combined to create a topographic surface used to 
direct water movement in the swamp hydrology model, and the peat thickness surface 
was used in comparisons of vegetation community types, fire history, and peat 
characterizations. 
Collection of Point Elevation Data 

During October 1991 -March 1993 permanent survey benchmarks were 
established at 86 locations within the Okefenokee National Wildlife Refuge and its 
perimeter. At each site a 3.2 cm diameter galvanized pipe was driven through the water, 
peat, and/or sand surfaces; one end of the pipe was buried at least 1 m into the sand, and 
the other extended 1-2 m above the water surface. Each pipe was topped with a 
galvanized cap onto which a 2.5 cm stainless steel bolt and nut had been welded. The 
top surface of the welded nut served as the reference point for measuring water depth, 
peat surface elevation, and depth to the sand basement below the peat surface; the bolt 
was the attachment site for a GPS antenna. A rebar probe was driven through the peat to 
the sand surface to estimate peat thickness, and a meter stick and tape measure were used 
to measure water depth and distance from the water surface to the reference point. The 
elevation relative to mean sea level of the top surface of the reference nut was estimated 
in GPS surveys conducted with assistance from U.S. Fish and Wildlife Service 
professional surveyors during October 1991 -March 1993. Differences between the 



151 
reference point elevation and the distance to the peat and sand surfaces provided 
estimates of the peat and sand surface elevations (Table 2-21). Control data were 
collected at 20 additional benchmarks in the swamp perimeter (Table 2-21). Absolute 
elevation for any surveyed benchmark was within 10 cm of first order mean sea level 
datum (NGVD 29). Elevation of any point relative to the nearest control point 
referenced in the survey was 6 cm. Between any 2 adjacent points within the same 
survey network, the error was < 3 cm. Several of the surveyed benchmarks also served 
as support poles for the water level recorder platforms; these benchmarks were the 
elevations to which the recorders were referenced. Stevens chart recorders, water depth 
staffs, and digital recorders not installed at permanent benchmarks were referenced to a 
benchmark located within 500 m of the recorder station using a laser transit and level. 
Additional points were surveyed among the GPS benchmarks during November 
1991 -April 1994 to improve the spatial resolution of the topographic database. A laser 
transit and level referenced to nearby GPS benchmarks were used to survey points in 
Chesser, Grand, and Durdin Prairies, and the sill dike. A "flatpool" survey was also 
conducted during high water periods to improve data resolution; the water surface was 
determined flat by reference to nearby benchmarks. At each "flatpool" point, percent 
cover of vegetation types in a 30x30 m plot was estimated independently by 2 observers 
and averaged. Water depth measurements were made at 3 locations within each 
vegetation type, and the average depths were weighted by the averaged vegetation 
percent coverages to estimate a site water depth, which was referenced to the nearest 
benchmark to estimate peat surface elevation. Three depth measurements to the peat and 



152 



E 


■ 



■ 


c J 
































1 


o en 




</"> 


so 


ov 


r- 


l> 


rr 


o 


m 


>o 


so 


■<*■ 


<n 


, ^- 


■* 


s 


1 ^ 




m 


■^r 


m 


o 


o 


Tj- 


CN 


00 


r- 


SO 


■* 


ov 


t^- 






S 


«* 


rn 


■<*■' 


Tf 


■*' 


rn 


CN 


CN 


CN 


rn 


rn 


rr' 


Tf 


1 


1 
1 


- s 




m 


rn 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


u 

60 






































































a 




i S 


3 






























u 




v e 


ce 




■* 


00 


o 


r^ 


CN 


00 


«T) 


«n 


<r> 


rf 


m 


'— 


so 


04 


■ 


!! i 


1 




CN 


so 


Vi 


CN 


■* 


m 


<n 


*r> 


«r> 


, ^- 


<n 


m 


•* 


£ 


£ 


< 




SO* 


</-> 


vd 


vo 


so 


so 


so 


vb 


so 


so" 


sb 


vb 


sb 




9 «* 


E 




m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


2 






** 






























*^- 




































s 






Cy 






























I 




e 


«- 


v> 


r- 


fN 


oo 


CN 


00 


so 


•^ 


Ov 


00 


r-^ 


CN 


so 


o 




> e 


1 


ov 


oo 


00 


SO 


Ov 


so 


»— H 


o 


CN 


t^ 


m 


00 


m 


o 




Si 


o 


1—1 


»— 


o 


>0 


o 


[*• 


CN 


>/"> 


i— 


«n 


so 


r» 


CN 







V 


^^ 


o 


00 


00 


VO 


so 


* 


** 


cn 


•<*■ 


m 


CN 


CN 


-ar 


•« 




5J 


o 


o 


o 


ov 


e> 


O 


Ov 


Ov 


Ov 


Ov 


Ov 


Ov 


Ov 


o 


o 


ca 




o 


5 


TT 


m 


m 


cn 


m 


m 


m 


m 


m 


m 


cn 


•<r 


>^- 


Z 






u 


cn 


en 


rn 


rn 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


a. 




































1 






u 






























00 




5,3 


1 


«o 


00 


Tj- 


vo 


m 


so 


p4 


m 


<o 


r<t 


Ov 


CN 


O 


so 




1 


00 


CN 


oo 


w-i 


t^ 


^f 


CN 


Ov 


o 


o 


00 


O 


m 


o 


1 




■5 


00 


*■ 


m 


VO 


m 


o 


Ov 


CN 


vo 


— 


V~i 


TT 


^r 


*n 




E 


o 


S 


vo 


r^ 


vo 


yn 


, ^- 


m 


m 


CN 


CN 


O 


so 


00 




© 


o 


00 


00 


oo 


00 


00 


oo 


oo 


00 


00 


00 


00 


00 


00 


— — 




© 

c 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


kefe 
ch si 






































































O R 




































a> 


E 


- e 


J 






























•9 ea 


■ 


3 .O 


GQ 


</-> 


m 


f— i 


o 


O 


f— i 


>o 


h« 


O 


O 


Tl- 


^tr 


^~ 


pot 


*-» 


1 £ 9 


2 


so 


—* 


o 


00 


CN 


*r 


CN 


v^ 


00 


Ov 


CN 


00 


t> 


r> 


.9 3 






< 


t-»' 


00 


K 


K 


00 


K 


oo" 


K 


K 


K 


00 


i< 


t< 


S 




e 


1 


r*i 


rn 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


•c 5 




































en ^3, 




































*— ~^m*S 




































stab! 
evel 




fa 
































1) — i 




E 
































vi ea 




s 
































w on 




Z 
































B ca 




E 
CO 




m 


— 


rs 


t» 


00 


m 


so 
m 


^t 


m 


vo 


m 


00 

cn 


>Tt 


so 


M § 




E 
































c c 




































t> t) 




e 




































& 
































o -S 




































» ■ 




































tion 
ions 




































w a 




e 
































o ?» 




e 
































o > 




•J5 
































— o> 




X 
































TJ 4> 




D 

2 

E 

« 

E 
































tions an 
surface 








'C 

CB 


E 

T3 
C 

o 
o 

B 




ea 




















leva 
and 




u 

B 

ea 






C 
Oh 
La 




(U 


2 
a- 




















e2-21. E 
peat and s 




<— 
© 

B 

e 






CD 
vx 

<*) 

U 


'C 

C 

Oh 


■a 

4> 


6 

e/j 
an 

D 

JB 




'B 
| 


■g 
| 


■g 
| 


;g 

E 


'C 

E 


o 

;g 


u 

;g 

E 


'C 

2 








< 


O 




> 
O 

i 


U 


•a 


(X 


Oh 


CL, 


a. 

T3 




"O 


Oh 


0- 








K 


E 

o 




■*-» 


c 


c 


C 


C 


c 


c 


c 


"u 


"o 


•S "o 













3 

o 


2 


S 


ca 

E 


ca 

E 


2 


ca 

E 


2 


N 


N 


5 a 








00 


2 


u 


00 


oo 


O 


O 


a 


o 


a 


o 


O 


s 


i 



153 



•o 

cu 
3 
C 
hO 

c 
o 
u 



fN 

i 

CN 

« 

H 



V 


































U -— - 


































tin 




































<n 


00 


* 


^-h 


o 


o 


00 


'O 


CN 


«— • 


r~ 


^ 


<n 


<n 


rr 


a 5 ? 
2 1* 




fN 


p 


>— ' 


in 


so 


ON 


r~; 


00 


s 


o 


sq 


«/-) 


•-H 


On 


ON 




rt 


«ri 


■* 


<n 


■<fr 


CN 


rn 


rsi 


en 


rn 


fN 


rn 


Tf' 


pal 


rf 


la i 




cr\ 


en 


en 


en 


en 


en 


en 


en 


en 


r<^ 


en 


en 


en 


en 


en 


en 




































^^ 
































« „ e 


_ 
































h* Q -ns 


c/2 


t> 


On 


■■N 


ON 


m 


o 


oo 


*H4 


■* 


^—< 


fN 


>o 


ON 


so 


00 


1 <r " 

a. Is e 


£ 


•O 


m 


en 


m 


■* 


so 


~* 


fS 


— 


fN 


fN 


m 


<n 


SO 


<n 


< 


so 


so 


so 


so 


so 


«ri 


so 


so 


sb 


SO 


SO 


so 


sd 


SO 


t> 


3 0* 


5 


in 


m 


m 


m 


en 


en 


m 


m 


m 


m 


en 


en 


m 


en 


en 


■ 


a 


<*n 


o 


r^- 




SO 


ON 


en 




en 


o 


r- 


oo 


in 


SO 


so 


> 5 


1 


so 


fN 


ON 


^H 


o 


Tf 


so 


(N 


— 


en 


so 


so 


t> 




* 


si 


OS 


00 


en 


t> 


ON 


fN 


00 


O 


oo 


ON 


o 


SO 


■* 


r> 


o 


■5 


SO 


^"H 


O 


CO 


en 


o 


r^ 


en 


ON 


t^ 


en 


o 


rf 


«r» 


en 


5 J 


© 


O 


o 


O 


C 


o 


1— « 


o 


«-^ 


o 


o 










fN 


O 


t 


5 


* 


Tf 


■<t 


Tl" 


3 


Tf 


■f 


5 


<* 


* 


«* 


TT 


^t 




Q 


en 


en 


en 


en 


en 


en 


en 


en 


m 


ro 


en 


en 


m 


rn 


en 


z| 

C 9 

3 .3 


1 


v> 


•* 


o 




•<r 


cm 


en 


On 


en 


m 


en 


<n 


fN 


r> 


ON 


i 


00 


so 


CN 


SO 


m 


cn 


00 


<r> 


so 


»r> 


m 


m 


so 


m 


m 


■5 


r- 


■* 


On 


* 


•■ H 


o 


00 


O 


»o 


so 


>n 


o 


<o 


m 


en 


- 


oo 


■* 


T 


fN 


CN 


00 


CN 


o 


m 


ro 


fN 


fN 


Tj- 


SO 


ON 


o 


oo 


oo 


00 


00 


oo 


r> 


00 


oo 


oo 


00 


00 


oo 


00 


oo 


00 


o 

u 


m 


m 


en 


m 


m 


en 


en 


en 


m 


rn 


en 


m 


en 


m 


en 


jl 


































1 9 S 


J 
































08 3 © 

J. w ■ 

S: — £ 


1/3 


fN 


— 


NO 


CN 


CN 


<o 


CN 


so 


</-> 


00 


o 


O 


so 


so 


en 


2 


«r> 


—" 


O 


en 


ON 


•"" 


SO 


fN 


>n 


«/-> 


fN 


rr 


r>> 


00 


■<r 


< 


00 


00 


00 


e> 


f»" 


00 


r> 


K 


f* 


S 


t> 


S 


i< 


K 


00 


Be — 


£ 


en 


m 


en 


en 


en 


en 


en 


en 


ro 


rn 


en 


en 


m 


en 


en 


k 


































1 


































E 


































3 


































z 


































E 

99 




SO 

00 


On 
en 


o 


•* 


CN 


SO 

SO 


r^ 


00 


On 


o 


— 


n 


fN 


m 


H— t 

en 


E 


































J 


































D 


































I 


































] 


































a 


































e 


































••e 


































03 


































y 


































.3 

E 




CU 






























09 




'3 






























E 




E 










2\ 






















£ 










e 






















■ 




.22 










§ 




















■ 




-*-* 










a 














cu 


cu 




03 




5/3 










£ 














•c 


'C 


cu 
•| 

'3 
c 

Oh 

c 

1 


<— 

e 

■ 
e 

f 
M 




5 

u 

o 

-= 

H-» 
3 


'C 

I 

J* 
O 


cu 

•c 

I 

0- 

M 
o 


CO 

CQ 
cu 

1 


CQ 
I 


3 
C 

W 
+^ 

6 


C 

| 
£ 

a 


— 

B 

'S 

i— 

ex 

CU 

</l 


'C 

'5 

E 

Oh 
CU 

en 

cS 


•c 

I 

CU 

cu 
<n 
03 


cu 
1 

Oh 
CU 

« 


'C 
1 

Oh 

en 

ca 


| 

o 

t 

cu 


I 

Oh 

b 

o 

1 

cu 
H 






o 


3 


3 


O 


O 


j= 


-c 


x: 


-c 


JC 


JZ 






oo 


CQ 


CQ 


U 


U 


u 


u 


U 


u 


u 


U 


H 


Q 



154 



£ = - 
■o V _ 



« 8 J a 

C li e < 

■"aw 

riE g | 



- f 
"** .2 a 

Hut 
. © I 

— — ~ 



Ml 



*^a 

5 £ .2 Z 

a 1 §■£ 
j H B 



en 

i 

I 



E 

a 
Z 

- 

s 



a 



00 






CO 



5 

CO 



m 

oo 

en 



fN 

00 

co 



o 

fN 



o 
co 

fN 

m 



os 

00 
CN 

co 



SO 

co 
co 



n- 

CN 

co 
m 



so 
00 

fN 



t-» 

fN 

m 



fN 

co 



fN 
fN 



CO 
On 



fN — 
co co 



so 
m 



OS 

so 
co 



oo 

Os 

so 
CO 



Os 



co 



fN 
fN 

co 



O 
SO 

co' 

co 



fN 

co 



in 

co 



Os 
co 



co 
co 

co 



fN 

iri 

CO 



IT) 

CO 



co 
© 

CO 



co 

CO 



so 


^H 


Os 


— 


fN 


o 


^— 


Os 


fN 


— 


■<r 


tT 


co 


CO 



Os 
fN 

m 

SO 

co 



o 
»n 
co 



so 
in 
fN 
«/-) 

fN 
CO 



fN 
fN 



CO 



Os tT 

r^ o 

o o 

fN fN 



co 



co 



Os 
fN 

SO 

co 



Tf CO 

oo t-» 
ir> in 



CO 



co 



oo 

Os 

co 



fN 
CO 



o 
co 
Os 

SO 

CO 



Os 



oo 



co 



CO 


fN 


Tf 


co 


*r\ 


SO 


© 


Os 


Os 


00 


co 


CO 



00 
CO 
Os 
00 

co 



VS. 

fN 

O 

o 

Os 

CO 



fN 

00 
CO 

OS 

CO 



Os 

fN 

00 

so 

CO 



o 

fN 

r- 

co 



SO 

o 

fN 
CO 



CO 
fN 

co 
co 



00 

*? 

CO 



o 

fN 
Os 
«n 
r^ 

co 



fN 
fN 

SO 

co 



o 
o 

fN 

00 

r-« 

CO 



fN 

r^ 

00 
SO 

co 



fN 

00 
CO 

CO 



Os 
fN 
00 
CO 



T}- w, 00 

rs) — | — 

00 OS 00 

co co co 



so 
fN 

OS 
CO 



SO —< 
Os co 



CO 



so 

CO 



OS 
«n 

so 

CO 



o 

fN 

SO 
CO 



Os 

CO 

sb 
co 



SO 

SO 

so 

CO 



Os 
co 

r-' 

CO 



Os 



00 

o 

r-.' K 

co co 



so 

sd 
co 



•a 

o 

3 

a 

«>» 

c 
o 
o 



fN 
i 

fN 



e 
■- 
m 
w 
o 

- 

- 

E 



a 



a 
e 

i 

as 



fN 


co 


■<* 


■* 


o 


fN 


t^ 


_ 


o 


Os 


"4- 


oo 


SO 


»0 


r- 


CO 


CO 


«r> 


co 


co 


V) 


SO 


«n 


«n 


rf 


Tf 


"3- 


* 


, ^- 


rf 



'C 

I 

C 

c 
-5 

3 

Q 



'B 

'3 

i- 

c 

i 

3 

Q 



u. 

o 
o 
<u 

oi 

u 
;£ 

'S 

c 
•5 

B 
3 

Q 



M 

«J 

-J 

c 

O 

o 

— 
a 

X 
o 

■♦-» 
3 
O 



60 

c 
c 

ea 

c 

5 



•S 
s 

V- 

O 



o 
T3 

u. 
O 

o 

co 



•a 

c 
W 

o 



ffl 





_*: 


t> 


CO 


^ 


_J 


CO 




J 


va 


'S, 


U 


>-, 


c 




c 






OQ 


2 


-— 


u- 


o 


o 


■a 


T3 


c 


C 


W 


W 


x: 


j= 


t: 




o 

2 


o 



0) 
CO 
CA 

'c 

c 



o 

T3 

C 
W 

— 

t: 
o 

z 



c 

I 

en 

O 



-2 

o 

o 

o2 

'C 
J 

O 



(U 

■£ 

'« 

t< 

o 



;g 

'3 

t> 

Oh 
CO 

O 



o 

"c3 

— 

CO 

>. 
O 






155 



•a 

3 
3 

I 
O 



i 

CN 

1 



■ 




































fl 




































•s 


B - 


































■ c/> 




o 


CN 


CN 


o 


o 


■^ > 


00 


U-) 


vo 


vo 


o 


^— 


*n 


00 


»^ 


S 

as 


« 2 




o 


00 


00 


o 


CN 


fS 


ON 


<— 1 


r-» 


vq 


Ov 


r~ 


p 


r>- 


VO 




r*i 


co 


co 


en 


m" 


rn 


(N 


«n 


CO 


•cS 


<^' 


vS 


«ri 


■^ : 


Tt 


1 

I 


- - 

B 1 




t*l 


m 


m 


co 


m 


ro 


CO 


CO 


<o 


co 


co 


co 


co 


CO 


CO 


09 








































_ i _ v 


































"„ ■ 


-J 


































« e 

w "IE 


« 


CN 


»— 4 


r-~ 


en 


v> 


r» 


CO 




r- 






«n 


vo 


CN 


o 


as 


■ E 


5 


vo 


o 


<— i 


o 


o 


ro 


>— 1 




CO 






CO 


CO 


CN 




£ 


< 


r*j 


** 


■*' 


Tf' 


<g? 


■* 


«f 




Tf 






p* 


S 


K 


s 




2 1< 


E 


co 


m 


en 


co 


co 


ro 


CO 




CO 






<o 


co 


<o 


co 




e 


£ 


</-> 


«*j 


t» 


«n 


co 


•* 


fN 


O 


vo 


CO 


■* 


00 


co 


o 






> 5 


re 
e 


oo 


00 


«n 


CN 


<N 


■* 


CO 


00 


co 


Ov 


CN 


CN 


00 


VO 


00 




s 1 


TJ" 


* 


On 


Ov 


00 


VO 


K 


00 


— 


<r> 


■<fr 


o 


VO 


to 


— 




1 


Ov 


00 


F> 


4— i 


CN 


"3- 


o 


f» 


«r> 


On 




OV 


o 


CN 


CN 




5 J 


v. 


© 


o 


o 


»— * 


4—1 


4-H 


4—1 


t-H 


^m 


O 


^-4 


CN 


co 


CO 


co 




o 


5 


Tf 


rf 


5 


■t 


■* 


5 


■* 


<* 


3 


rr 


^l- 


■<d- 


<<*■ 


•<* 









rn 


co 


m 


rn 


co 


m 


co 


co 


co 


<o 


co 


co 


co 


co 


co 




5 3 


1 

re 


■<i- 


O 


r- 


co 




vo 


CO 


en 




«r> 


VO 


vo 


CN 




00 




i 


00 


00 


o 


co 


I— H 


Ov 


Os 


Ov 


r^ 


o 


<n 


Ov 


co 


co 


r- 




•S 


vo 


r^ 


</-> 


■* 


'S- 


■* 


CN 


<n 


CN 


On 


, ^- 


O 




o 






c 


vo 


tT 


•t 


■t 


Tj" 


rr 


VO 


■* 


Tf 


rf 


•<*• 


VO 


VO 


<n 


o 




1 

o 

u 


VO 


VO 


vO 


VO 


VO 


VO 


VO 


vo 


vo 


VO 


vo 


00 


00 


00 


00 




co 


«*J 


co 


CO 


co 


m 


CO 


co 


CO 


CO 


co 


co 


CO 


CO 


co 


ji 




































E 


C B 


2 
































■ 


a ■ 


in 


r* 


r» 


<n 


VO 


00 


co 


00 


co 


Ov 


r-» 


vo 


00 


CM 


VO 


00 


1 £ « 


2 


*■ j 


rn 


00 


* 


■<*- 


00 


p- 


vo 


t*- 


«r> 


vO 


■* 


r- 


<r> 


Tt 


fl 




< 


<n 


«r> 


</"•>' 


<o 


<o 


«/S 


vi 


vd 


u-> 


vd 


vb 


00 


00 


00 


00 


e 
I 
CO 


£ 


CO 


m 


co 


co 


en 


co 


co 


co 


co 


co 


CO 


CO 


co 


CO 


CO 




.8 




































E 




































9 




































z 




































E 

re 




00 


r~ 


co 


tT 


o 


■_i 


Ov 


CN 


<n 


m 


vo 


Ov 


00 


r^ 


vo 






<ri 


iri 


VO 


vO 


VO 


VO 


>r> 


vO 


VO 


<o 


*rt 


CN 


CN 


CN 


CN 




E 




































.= 




































B 




































B 




































B 




































CO 




































e 




































e 


















ha 

1> 


















re 


















"O 


















o 




H 














E 


















■3 




4* 
1 






1 


1 


1 


1 


O 

o 


1 
















■ 




o 






H 


H 


H 


H 


C* 


H 
















E 




Q 






-* 


M 


-ui 


-^ 


_^ 


-^ 


ss 


JS 












JS 


B 

=1 




i* 






o 

o 


u 

o 


u 

O 


o 

o 


u 

o 


o 

o 


co 


co 














6 






£ 


£ 


E 


E 


E 


E 


Lj 

p 


P 








-* 






> 






P 


E 


E 


£ 


E 




> 


> 


V3 






8 




O 

c 




u 






ffi 


X 


a 


a 


CO 


1 


5 




<D 


<U 


E 

i 

X 

~3 




•a 

■ 

0£ 




i 


< 


CO 

1 


C/J 

fl 

i 


"c 
> 

at 


a 
> 


«5 

> 

CO 


fl 

> 

CO 


fl 
p 

> 




i 


c 


3 




i 

o 








3 


, M 


m3 


E 


S 


C 


E 


E 


E 


3 


3 


o 


O 


J& 








1/3 


CO 


CO 


u 


U 


U 


o 


o 


U 


CO 


CO 


Q 


O 


S 



156 



-o 

3 
_C 

c 
o 
o 



rN 
■ 

rs 

u 

-O 

a 
H 



V 
































U .— s 
































,2 | -J 
Tl .2 <Z> 
































o 


m 


O 


SO 


m 


fN 


rN 


r-~ 


■* 


«r> 


rr 


^* 


•*t 


00 


fN 


and Su 
Elevati 
(m AM 


* 


fN 


■* 


</i 


so 


r-~ 


00 


r^ 


o 


Os 


o 


r~- 


00 


CN 


!>• 


•**■■ 


rf 


■*' 


"3- 


•<r 


* 


■* 


>ri 


•*' 


rn 


S 


Tf 


vri 


00 


K 


en 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


en 
































































V 1 - 1 
































.Sin 


■<3- 


so 


so 


^. 


■<r 


00 


SO 


rl- 








«— H 


rM 






Ills 


os 


so 


00 


SO 


so 


rr 


<o 


V 








— 


r^ 






so" 


so 


so 


so 


sd 


SO 


so 


so 








sb 


so 






3 ■ . 


en 


m 


m 


en 


m 


m 


m 


ro 








m 


m 






TMY 

Dcation 

Drdinate 


tj- 


■* 


o 


fN 


-3- 


rN 


o 


'i- 


r^ 


so 


o 


Os 


m 


fN 


m 


»— • 


fN 


m 


O 


^-< 


r^- 


^-t 


■* 


o 


rr 


>o 


m 


so 




00 


*■■* 


On 


o 


^^ 


•* 


^■M 


o 


rN 


m 


r^ 


r-« 


r- 


OO 


00 


os 


m 


^* 


*"H 


o 


Os 


rN 


00 


r^ 


00 


i— i 


«r> 


o 


fN 




o 


m 


m 


m 


m 


CN 


m 


rN 


rN 


Os 


Os 


rN 


o 


OS 


Os 




— _; ° 


*■ 


•t 


rt- 


TT 


rf 


■* 


rf 


■^r 


m 


m 


ti- 


"3- 


rn 


en 


TT 


U 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


vIX 

ation 
dinate 


r» 


o 


Os 


•<a- 


rn 


</1 




(N 


■* 




O 


r- 


o 


O 


o 


o 


>n 


OS 


^H 


m 


m 


V~l 


rN 


00 


* 


m 


Tf 


rn 


so 


•n 


oo 


r^ 


<o 


*r> 


r^ 


so 


r^ 


rs 


so 


o 


•^• 


r^ 


o 


m 


•<a- 


UTl 

Loc; 

Coor 


oo 


SO 


r» 


t^ 


t^- 


Os 


so 


r-- 


r^ 


so 


o 


o 


o 


t-~ 


CN 


t> 


r» 


r» 


r» 


r» 


r^ 


r- 


r<- 


*r> 


v-i 


so 


r^ 


00 


r» 


r* 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


-z *» 
































- C C — 
































ichma 
p(Nu 
evatio 
AMS 


rn 


so 


m 


so 


o 


CN 


rN 


Os 


</-> 


so 


rn 


r^ 


so 


m 


r- 


</"> 


T 


~ 


r^ 


00 


rN 


t> 


00 


rn 


00 


rN 


so 


m 


m 


r- 


00 


00 


00 


t> 


K 


ob 


s 


K 


sd 


V) 


00 


00 


oo 


00 


r».' 


- c — — 


rn 


rn 


r*"i 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


m 


Lt 
































J 
































s 
































a 
































Z 
































E 
as 


«/-> 


^J- 


fN 


m 


,— 


m 


o 


Os 


o 


>o 


00 


Os 


_ 


^_ 


rN 


CN 


fN 


fN 


«n 


rs 


CN 


rs 


^. 


r*. 


00 


oo 


so 


00 


t-- 


r-- 


E 
































J| 
































<j 
































i 
































1 
































■ 
































© 
































•c 
































■ 


















u. 














Q 


















B 














© 








h 










| 














■ 








^ 
^ 










8 

o 






1 








E 
J 










a: 






n 


A 


o< 






1 








e 












<u 


^ 


a 




^s 





u 


T3 

J 
co 

y 




1 

e 

e 
© 


'C 
J 

Oh 
00 


■ 22 

'C 

2 

0h 
00 


;g 

'3 

B 

Oh 

00 


'C 

'3 
c 

0h 
00 


u 

■c 

J 

Oh 
00 


;g 
s 

Oh 
00 


1 

c 
o 

Oh 

G 

c 
c 


I 

•o 

c 
o 

0- 

La 

D 


U 

L) 

i 

6 

p 


U 


u 

c 

U 

u 
D 

c 


K 
_u 
[C 

"3 

Lh 
Oh 


'3 

E 

Oh 

CO 


T3 

C 


■ 


C 

"5. 




c 
"o. 




"S. 




CO 

p 

D 

a. 




y 

c 




CO 




C3 


A 


CO 


« 


R> 


a 




»9 


>^ 


3 


o 


P3 


P— 1 




CO 


CO 


CO 


co 


CO 


CO 


Q 


Q 


CO 


U 


CO 


X 


OQ 


s 


5 



157 



-o 

3 

a 

C 

o 



CN 

i 

CN 

V 

3 
H 



41 
































u .— . 
































£ = - 

fc .2 E9 
































en 


00 


ON 


>o 


r-- 


On 


O 


CN 


00 


00 


>o 


en 


iri 


en 


00 


s x 5 


<r> 


00 


</■} 


o 


■«i- 


CN 


so 


>-< 


o 


CN 


en 


■* 


r» 


K 


t— * 


00 


f* 


so 


00 


so 


00 


d 


00 


d 


00 


<ri 


so 


■>* 


</-! 


«ri 


73 V 

gS E 

BO 


en 


m 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


^> 


































- B _ 
































Peat 
urface 
evatio 

AMS 
































m EE 
































c 1 


en 


<T) 


o 




r-- 


ro 


o 


r-- 


so 


ON 


00 


CN 


tt 


rf 




> J J 


SO 


<o 


</-> 


fN 


CN 


»-H 


On 


NO 


en 


i—i 


en 


so 




so 


On 


(•» 


SO 


*-^ 


<n 


fN 


r^ 


00 


<r> 


h- 


en 




O 


^d- 


en 


On 


u^ 


r» 


00 


en 


«— H 


oo 


»— H 


■* 


en 


so 


en 


r- 


CN 


TT 


O 


c o e 
3 _j c 


^^ 


fN 


«— 1 


o 


on 


00 


*— 1 


o 


oo 


en 


^m 


r» 


00 


00 


O 


Tf 


•<* 


T 


^t- 


en 


en 


■^r 


Tt- 


en 


•* 


Tf 


en 


en 


en 


TJ- 


C 


m 


rn 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


UTMX 

Location 
Coordinate 


Tt 


•*r 


</-> 


so 


IT) 


so 




en 


<o 


IT) 




en 


r- 


en 


O 


ON 


CN 


en 


o 


"3- 


en 


»-^ 


*— i 


00 


CN 


♦■^ 


o 


o 


IT) 


On 


r^ 


<N 


On 


t-» 


t-» 


*— H 


CN 


SO 


^— 


00 


ON 


o 


10 


o 


On 


o 


00 


© 


00 


i—( 


1-H 


V> 


CN 


ON 


CM 


CN 


en 


en 


en 




oo 


SO 


(^ 


r^ 


so 


00 


«0 


r«- 


r- 


r^ 


SO 


00 


oo 


00 


On 


en 


en 


en 


en 


en 


en 


cn 


en 


en 


en 


en 


en 


en 


en 


en 


-i: *-» 
































- « c — 
































china 
p(Nu 
:vatio 
AMS 


<n 


rr 


*n 


00 


00 


*• 


CN 


00 


en 


so 


CN 


en 


<o 


en 


00 


«r> 


o 


NO 


^ h 


m 


■* 


r^- 


<— ■ 


<o 


so 


CN 


VN 


00 


00 




00 


r*» 


so' 


00 


so" 


oo 


d 


oo 


r-' 


On 


S 


so 


rf 


•ri 


iri 


& H H 1 


{*! 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


en 


Ti- 


k 
































J 

E 


























CN 






























o 


en 


es 
























^^ 


o 


o 


o 


3 
























o 




»— 1 


T"^ 


z 


en 


TT 


u^ 


SO 


r-- 


00 


ON 


o 


CN 


en 


Tfr 


o 


1 


1 

en 

u 


CN 

a, 
U 


la 
■ 


S 


r^ 


t-- 


r^ 


r-~ 


r- 


r-~ 


00 


00 


oo 


00 


■4 


Z 


E 
w 
























5 


2 


1 


a 


I 

4) 

■ 


























o 


o 


■ 

e 
•■c 

s 
























1 






T3 

Cm 


e 
























#s 

c 








M 
E 
























o 






o 


1 




T3 




















o 




T3 

o 


o 


E 




C 

a 




















§ 




"aS 


D 
■ 

J 

e 
■ 


T3 

C 


"en 

o 

c 

2 


T3 

C 

"c75 


c 

en 


T3 
C 


-o 

c 

-2 

"v] 


1— H 


c 


00 

T3 


u 

o 

E 
E 


T3 

c 


00 
1 

ON 


T3 

a 

3 
O 

2 


a 

£ 

u 


3C 
.2 

E 


f 




CQ 


VJ 


O 


en 


^.^ 


cn 


~an 


«j 


>-> 


en 


en 


o 


VI 


CO 

E 

o 


"a 

■ — 

c 


o 


B 


ID 


"o 


1> 


3 

en 

c 




t 


O 




U 






c 


00 

3 

m 




o 


O 

(5 


C 

o 

X 


2 


CO 

CQ 


c 


00 




2 

o 

00 


E 

03 





158 



■ 
tin 

s'S5 

2 I* 

«5 



« -S « g 









ON 






o 

NO 

m 



fN 

ON 



-3" 



NO 

fN 

00 






_ — o 

•* ^ ^ 

00 On On 

m m m 









-3- 
t-- 



NO 
NO 

l> ON 



■s E 

S ■ -p 

E" c o 

; c 

U 




V3 



« 3 O „ 

w E 



<s 



o 



fa 

E 

a 
Z 

- 

OS 

E 

W 

I 



00 
ON 
NO 
V) 

O 



fN 

fN 



NO 
NO 

T 
fN 

m 



NO 

ON 
fN 

m 






ON 
NO 

m 



oo 
oo 
<n 
en 



NO 

■^- 

o 

m 



00 
00 
CN 

m 



o 

fN 
m 









ON 

r-- 

oo 

m 



r- 

O 

fN 

00 

m 
en 



o 

00 
ON 

CI 



00 

O 

fN 

ON 

m 



NO 
00 

CN 

ON 

m 



o 

ON 



00 

fN 
ON 



NO 

r- 

fN 

00 



V) 



00 



oo 
m 
m 
o 

00 

m 



fN 



ON 
NO 



NO 
NO 



ON 
NO 

m 



ON 

NO 

00 
ON 

vo 



r^NOrn«0'^-NOf < ^ — >— om 
ONOo\©ONv>m'«3 - CNi</">ioNC 
wi rt «o «r» rt- oo od oo' on on v-> 



— o 



o 

o 



T3 

3 

c 
o 
a 



fN 



e 
e 
- 

SB 
O 

o 

- 



s 

E 

-c 
w 

B 



B 
c 

'St 

o 



00 
< 

Q 



00 



U 






Si 



00 

o 
o 

I 

H 

fN 



u 



ON 

o 
o 



< 



0-. c/3 

u s 

g i 

s 8 



en 

o 
o 

I 

E 



2 Z S 

m o — « 



oo 

ON 

o 



00 

u 

00 



S o 



U 

o 



1/3 

•a 

O 

K 

CO 



s 

3 

QQ 

-o 

c 

ra 



oo 
o 



c 
o 

on 
O 



p -1 



n 
O 



o 
N 



E £ 



On 
(N 

oo 

D 




o 

NO 

in 

NO 



oo 

NO 
NO 
NO 



rs 






o •* 

00 VO 
NO* l> 



ON 

m 



o 



oo 



e2 



00 



00 
fN 

00 
ON 

m 



vO 
(N 

ON 

00 



od 
r<1 



ON 

5 2 



p« d. c* y 

o ^ ^ 



fN 





D 




oo 






c 




O 




T3 


b 




00 




ca 
o 


>-, 




D 




2 




O 

H 

2 


c 




2- 




IX, 


a> 




C 


B 


>. 


V) 




r. 




•o 


B 




C3 


fi 


T3 


j= 




oo 


M 


UJ 


U 



































159 



!> 
3 

B 

c 
o 



i 

fN 

H 



■ 








B •*?■ 








£ = - 








t in 




VO 


ti- 


8*5 




fN 


VO 


— t < 




v> 


vi 


ja ■ 




m 


m 


n 










^^^ 






a. B 


_ 






Peat 
urface 
evatio 


< 






SB s 


£ 








a 


r- 


r*- 


1 


CM 




■5 


O 


00 


E s 


- 

o 




o 


3 j 


© 


rr 


•<r 







m 


rn 


X § 


1 


</■} 


00 


c 


V© 


^— 




'1 






o 


VO 


vo 


=> - 


o 




m 


cn 


^ „ 








•- ~ = 


J 






« s c 


t/j 


v© 


o 


nchm 
op(N 
levati 


1 

< 






£!-" 






k 








V 








A 






vO 


E 




■o 


O 


3 






^^ 


z 




© 


1 

O 


M 

E 
I 




— 
1 

H 


u 


E 
J 

B 




X 


a 


e 






o 


£ 
































c 
e 




00 


oo 


•■c 




u 


u 


■ 




D 


6 


3 




c 
2 


c 
c 






a 


a 


E 




Z 


Z 


■ 




3 


3 


E 




00 


00 


J 




Im 


t— 


i 




o 


o 


■ 




ca 


>-> 

CO 


e 




£ 


z 










■ 








« 










Q. 


Q. 


5c 




00 


00 


■ 




-C 


-c 






E 


3 






o 


O 






Z 


00 



B 

X) 
CO 

3 

O 



C 

o 

«-» 

CO 

o 

J5 



c 
u 
OQ 









160 
sand surface using the rebar probe were also made at the "flatpool" sites to use in peat 
thickness estimates. Peat elevation estimates were also made at all transect sampling 
points (see Chapter 6) and averaged to represent peat elevation in the transect area. To 
complete the point data in inaccessible regions of the swamp, point elevations were taken 
from 1966 USGS 7.5" 1:25,000 quadrangles after determining that the GPS survey 
elevation data agreed with the USGS elevation data in the swamp perimeter. These 
points supplemented GPS survey points on the large interior islands, or were outside the 
refuge perimeter (Figure 2-18). 
Surface Interpolation 

A topographic grid was created for the combined peat and sand surfaces (peat, or 
sand where no peat occurred) using ARCINFO-GRID's kriging procedure on the 
PEATELEV coverage item. Several algorithms and grid sizes were used, with the 
circular model and 500x500 m cell size resulting in the best semivariogram (Burroughs 
1986). The resultant surface was compared to the original data points to check the 
interpolation accuracy (Figure 2-19). A correction surface was added to the interpolated 
surface to adjust for interpolation errors, and the final grid was smoothed with a filter 
(5x5 cell, or 2500x2500 m, mean window) to eliminate pits and peaks in the estimated 
surface (Figure 2-20). 

The swamp peat thickness was calculated by differencing the sand elevations and 
peat surface elevations at each surveyed point; 19 additional peat thickness estimates 
reported by Cohen et al. (1984) were added to supplement the grid. Inaccessibility made 
collection of peat thickness data in the south-central region of the swamp impossible; 



161 




Figure 2-18. Locations surveyed and extracted from USGS 1994 1 :24,000 topographic 
maps for development of the Okefenokee Swamp topographic surface. 



CD (/) W 



TO 

c 
o 

1 



8§88£ 



= ^J5iS^2oio8o 
c75£J2wg< : ?9oo<r : Z 



«# 








sSSKSSSSS™.^ 




^*a^§|WX 




'-:.'-'.■.;.■.;.■;::;:■.-:■:■:::■:;■:■:-.:■:;■.■-::■:■-■ 









162 




pa 




§ 




6 




O 




s 




u 




€ 




c 




1 




1 




1 




V 




o 




g 




E 




ffe 




■-B 




a 




1 




E 




o 






£ 


•o 


£ 






CS 




E 


1* 


6 


c 


c 


o 


8 


2 


00 

a, 




03 


?r 


E 




>> 




JB 




a. 




03 




& 




o 




a 




o 




■4-t 




a 




E 




3 


8 






-e 




*- 




; *■• 




I C/5 




3 




w 




> « 




O _; 




i *■• ed 




v s 




D T3 

C/5 




3 a 


o 


J o 2 
■ o S 




is a 




le 5 




3.H 




1 ft U 




la e 




1 o o 




1 — 03 




n#d 




i) 3 




fc w 




O "O 




i 

-19. C 
rpolate 


^1 




<N B 




-, ■*-» 




g .C 




BO'S 




•~ c 




U- as 











163 




Figure 2-20. Peat and sand surface topography in Okefenokee Swamp. Darker areas are 
lower in elevation above mean sea level. 






164 
therefore, peat thicknesses in this area were estimated from those associated with 
vegetation types in other regions of the swamp. Five points were randomly selected in 
each vegetation type (except on large sand-based islands) in the south-central swamp. 
Peat thicknesses for each vegetation type in the remainder of the swamp were averaged 
and applied to the randomly selected points in the south-central swamp (Table 2-22). An 
estimated peat thickness grid was created by kriging this combined dataset with the 
circular model and 500x500 m cell size (Figure 2-21). A sand surface elevation grid was 
created by subtracting the estimated peat thickness grid from the original peat surface 
elevation grid (Figure 2-22). 

Topography Surface Description and Trends 

Topographic relief in the swamp is minimal. The swamp is a bowl-like 
depression in the landscape with the trend in ground surface elevations from 38.4 m at 
Kingfisher Landing in the Northeast to 33.0 m in the area where the Suwannee River 
exits the swamp in the West to 34.8 m at Ellicott's Mound in the Southeast near the St. 
Marys River outflow. Basement sand topography also follows this trend. Within the 
swamp are regional topographic highs on large sand-based islands and lows in large 
prairies and stream beds, ranging in elevation from 38.4 to 33.6 m AMSL. The prairies 
also contain local topographic highs on peat-based islands that may raise a meter above 
the surrounding inundated peat surface (Figure 2-23). This local topographic variation 
results in gradients of vegetation community distributions within the prairies; the forest 
matrix between the prairies has less topographic variation and a less diverse vegetation 



165 



Table 2-22. Peat thickness values used to estimate peat depth by vegetation type, to 
supplement the coverage of estimated peat depths where data gaps exist. 



Vegetation Type 


Number 
ofCeUs 


Area 
(ha) 


Mean 

Peat 

Thickness 

(m) 


Minimum 

Peat 
Thickness 

(m) 


Maximum 

Peat 
Thickness 

(m) 


Variance in 

Peat 

Thickness 

(m) 


Gum-Maple-Bays 


23 


575 


0.99 


0.45 


2.26 


0.32 


Water Lily 


63 


1575 


2.87 


0.51 


3.62 


0.32 


Gum-Bay- 
Cypress-Shrub 


476 


11900 


1.56 


0.37 


3.41 


0.62 


Mixed Wet Pine 


7 


175 


1.70 


1.58 


1.90 


0.01 


Sedges-Ferns- 
Water Lilies 


349 


8725 


1.90 


0.37 


3.67 


0.70 


Briar-Shrub 


124 


3100 


2.61 


0.69 


3.57 


0.46 


Open Water 


3 


75 


2.53 


2.13 


3.32 


0.31 


Bay-Shrub 


868 


21700 


1.82 


0.37 


3.63 


0.87 


Cypress-Gum- 
Shrub 


1217 


30425 


1.79 


0.37 


3.58 


0.88 


Loblolly Bay 


418 


10450 


1.61 


0.37 


3.20 


0.64 


Shrub 


174 


4350 


1.67 


0.38 


3.50 


0.75 


Dense Pine 


50 


1250 


0.90 


0.37 


3.24 


0.48 


Sparse Pine 


20 


500 


0.78 


0.37 


2.10 


0.43 


Mixed 

Upland/Wetland 

Shrubs 


9 


225 


0.75 


0.42 


1.24 


0.07 


Pine-Cypress- 
Hardwoods 


37 


925 


1.15 


0.38 


2.96 


0.49 



166 






c/5cCJ2J2'fDoc>o , «- : cMCNi 






*Sffl 



Ulj 








o 
2 

9 



8 



E 

I 

1 

O 
.c 

T3 



I 



JD 

I 
1 

8. 

■— 

o 



I 



8 



i 

E 






PL, 



167 




Figure 2-22. Estimated sand surface elevation above mean sea level under the surface 
peat in Okefenokee Swamp. 



168 







C/3 



n. 
E 

V 
<U 

o 
S 

u 

O 



o 
<u 

en 

e 
a* 



c 
o 

3 



o 

o 

U 



c 
u 

•3 

I, 

c 
5 

'£ 

> 



m 
i 

>- 



169 
composition (see Chapters 5 and 6). The regional gradients in topographic elevation 
direct water movement through the swamp, towards the Southwest and Southeast (see 
Chapter 3). Peat thickness is greatest in the prairies found primarily in the central and 
eastern swamp, most likely due to the ponding of water in these topographic lows which 
decreases decomposition of the accumulated peat. The peat surface is occasionally 
exposed during periods of extremely low rainfall, which occur every 20-30 years; 
oxidation during this exposure and removal of peat by widespread fires lower the surface 
elevation and result in greater inundation when normal precipitation resumes. 

Water flows through the swamp along natural and maintained rivers, creeks, and 
trails (Figure 2-24). The topographic surface in these drainages is terraced, so that the 
dendritic flow patterns visible on aerial photography and satellite imagery on the swamp 
represent local topographic highs or berms (Figure 2-25) over which water flows. 
Upstream from these berms are ponds and lakes (e.g., Dinner Pond, Big Water, Minnie's 
Lake, Billy's Lake, Cravens Lake) which crest over the berm during high water periods, 
and are impounded behind the berms when water levels drop. Beyond the berm summit, 
the elevation drops to the next "impounded" lake or pond in the stream bed. The sill 
probably acts as one of these berms in the Suwannee River drainage, so that during high 
water periods, water crests the sill gates and is impounded upstream to and possibly 
beyond the next highest elevation or berm at the southwest outlet of Billy's Lake. During 
drier periods when the river is confined to its banks, water is impounded to the northeast 
of Billy's Lake by this natural berm, beyond the area of impact of the sill berm (see 
Chapter 3). 



170 




Figure 2-24. Creeks, rivers, lakes, and canoe trails where surface water flow occurs in 
the Okefenokee Swamp. 



171 




Figure 2-25. Surface water drainage patterns and underlying topographic gradients in 
Okefenokee Swamp. 



172 
The regional topographic relief creates hydrologic basins and isolates the 
impounding affects of the sill to the western swamp. The ground surface connectivity of 
the large, sand-based islands (Floyd's Island to Billy's and Honey Islands and the Pocket; 
Strange Island to Blackjack and Mitchell's Islands; Moonshine and Soldiers Camp Island 
area) and intervening and intervening depressions are apparent in the swamp topographic 
map (Figure 2-20). Water in the central third of the swamp most likely does not drain to 
the St. Marys River or southwestern creeks even thought the primary gradient is in this 
direction, because these large islands impede flow. Cohen (1973b) believed that Floyd's, 
Grand, and Chase Prairie are persistent prairies, probably due to these sub-peat 
depressions being isolated by ridges (Davis 1987, Smedley 1968). There is some water 
movement along the Suwannee Canal to the west due to the overall topographic gradient 
in that direction; however, prior to the canal's construction, this water movement was 
probably restricted by this natural berm, and most water moving into the Suwannee River 
and westward probably originated west of the Floyd' s-Honey-Billy's Islands and Pocket 
chain (see Chapter 3). These topographic features determine the spatial hydrologic 
environment of the swamp, and subsequently influence distributions of vegetation 
communities (see Chapters 4 and 6). 
Satellite I magery Classification and Accuracy Assessment 

Classified satellite imagery provides a geographically referenced record of the 
vegetation community composition and distribution over a large area at a point in time. 
Depending on the satellite data scale and quality, imagery can be classified to provide 



173 
high resolution information of existing vegetation distributions. Lo and Watson (1994) 
classified Landsat thematic mapper data in mapping Okefenokee Swamp vegetation and 
determined that the 30 m data resolution was insufficient for making class distinctions in 
the patchy environment of the swamp. The spatial complexity of the swamp vegetation 
requires image data at a finer resolution. SPOT satellite imagery available from 
panchromatic (PAN; 10 m pixel size) and multispectral (XS; 20 m pixel size) scanners 
can be merged with transformation of the hue-saturation-intensity bands to provide data 
at 10 m resolution (Jensen 1986), a more suitable scale for mapping complex wetland 
vegetation communities. The classified maps can be compared with interpreted historic 
imagery or photography to assess occurrence and successional vegetation change 
(Silveira 1995). The classifications reported herein are used to document present 
vegetation distributions and change (see Chapter 4). The following accuracy assessment 
provides an index of map reliability and an indication of class confusions to consider in 
map interpretation and use. 
Image Preparation 

SPOT PAN and XS imagery were selected for the vegetation map. The most 
recent scene available providing growing season (March-October) vegetation and 
minimal (<10%) cloud cover was 1 1 May 1990. More recent imagery (through 1994) 
provided incomplete swamp coverage or contained interference from clouds. 

ERDAS version 7.5 and IMAGINE (version 8.2, ERDAS, Inc., Atlanta, GA 
30329) image processing software were used to prepare and classify the satellite imagery. 



174 
The PAN image was rectified to ground control points selected on 1966, USGS 1:24,000 
scale topographic maps (ERDAS 1995). The transformation matrix of the control points 
from the topographic quad sheet to the satellite image was generated using only ground 
control points with an error between the locations of less than one pixel (10 m). The file 
coordinates of the XS image were then rectified to the map coordinates of the PAN 
image, so that both images would be spatially registered to the same coordinate system. 
Nearest neighbor re-sampling was used to perform the rectification; this method does not 
corrupt the original band data so that subsequent image classification has not been 
compromised (Lillesand and Kiefer 1994). 

The complexity of the swamp vegetation requires high resolution data. 
Combination of the 3-band (green, red, and infrared wavelength reflectance) 20 m XS 
data with the single band (green-red wavelength reflectance) 10 m PAN data creates an 
enhanced image that uses the color information of the XS data with the spatial resolution 
of the PAN data. The bands are combined by re-sampling the XS data to 10 m 
resolution, transforming the XS data in red, green, and blue color space to hue, 
saturation, and intensity, and then substituting the XS intensity data with the 10 m PAN 
data (Lillesland and Kiefer 1994). Then the XS data are back-transformed to red, green, 
and blue color space, with the color intensity enhanced by the PAN data. The resultant 
merged image has XS 20 m spectral data at PAN 10 m spatial resolution in 10 m pixels 
(Figure 2-26). This conversion enhances edge features such as islands or ponds while 
retaining the spectral information, which facilitates identification of wetland vegetation 
composition. 



175 



Multispectral 



Red 20 Hue 20 



Green 20 Saturation 



20 



Blue 20 Intensity 20 



Panchromatic 



Hue 



10 



Saturation 



10 



•* Intensity 10 



Multispectral 



Red 20 


Hue 20 


Green 20 


Saturation^ 


Blue 20 


Intensity 10 



Figure 2-26. Merging 10 m pixel panchromatic and 20 m pixel multispectral imagery to 
create a multispectral image with 10 m pixel resolution. 



176 
A normalized difference vegetation index, 

[red - infrared/(red + infrared) * 0.5] * 100 

was also calculated and added to the merged file as a fourth data band to enhance the 

interpretation. This band emphasizes vegetation biomass, aiding differentiation from 

less-densely vegetated areas (Jensen 1986). 

Image Classification 

Training sites for the image classification were selected from an unsupervised 
classification of a 10 m resolution 1987 SPOT satellite image merged as described 
above. Approximately 100 large, single-class areas were selected from the classification 
for training (seed) and ground control sites. An additional 100 random sites evenly 
distributed among the four swamp quadrants were also selected as ground control points. 
Sites were visited during June 1994 by helicopter, and vegetation was identified in the 
2500m 2 (50x50 m) area below the helicopter, which hovered at an altitude of 50 m. Two 
observers independently determined the vegetation type and then compared their results 
to assign a vegetation class. Photographs were taken at each site for later reference. 
During June 1996, 46 additional sites were visited on the large islands to collect example 
points of island vegetation to use in classification improvements. 

Two PAN and XS satellite scenes were required to cover the entire swamp area. 
The images were captured as shifts north and south of the same scene, providing 
considerable overlap in ground coverage. The band data differed slightly among the 
scenes, requiring that they be matched and then classified. Matching the north and south 



177 
scenes was accomplished with histogram matching. This procedure compares the band 
data histograms and attempts to equalize them in the area to be matched (ERDAS 1995). 
The edges in the matched area were blended with feathering, and then the images 
wereclassified. The re-classified edge was then stitched back into the matched image, 
and the entire image was scanned with a 3x3 pixel majority scan to remove single pixel 
classes and residual match lines. 

An iterative methodology of selecting seed areas at the training sites on the 1990 
merged satellite image, examining the signature euclidean distances and photos of each 
site, and deleting or combining seed sites was used to identify the 36 classes to be used in 
the initial image classification. A supervised classification of the image with this 
signature set using the MENDIST algorithm, used due to non-normality in the image band 
data (ERDAS 1995), resulted in a classification with some class confusions. The 
classification was repeated by removing some signatures from the set, masking the image 
to include only specific areas, re-classifying the image portions, and stitching them back 
into the image. Class distinction was improved by combining signatures and eliminating 
classes. The band data differed slightly among the scenes, requiring that they be 
classified with two separate signature sets and then combined along the scene match line. 
To eliminate the match line, pixels in 4 classes (loblolly bay, gum-bay-cypress-shrub, 
gum-maple-bay, mature cypress-shrub) were masked from the match area and re- 
classified with a modified signature set using only band 1-3. The re-classified region was 
then stitched back into the composite map. The entire classified map was scanned with a 
3X3 pixel majority scan (ERDAS 1991) to remove single pixel classes and residual 



178 
match lines between the scenes. The classes in the final 22- class map were consolidated 
to produce the 17-, 13-, and 1 1 -class maps (Table 2-23). The 17-class map was also 
modified to include improvements to the upland island classification. The large, sand- 
based islands were removed from the 17-class map, and replaced with a revised 
classification of the subset area including 4 additional classes representing upland 
communities (dense pine, sparse pine, mixed upland-wetland shrub, and pine-cypress 
hardwoods). The resultant composite map of 21 swamp and upland island classes was 
used in all vegetation change and hydroperiod association analyses (Chapters 4, 5, and 
9). 
Image Classification Accuracy Assessment 

Ground-truthing data collected at 198 sites within the swamp wetland matrix and 
46 upland island sites were used to assess the accuracy of the classified maps (Table 2- 
24). Each site was located on the classified map and the area around it searched for the 
class of interest. Class occurrence was recorded within radii of 3-5, 10, and >10 pixels 
(corresponding to 30-50, 100, and >100 m) of the point location. These distances 
reflected the accuracy of the GPS point locations; differentially correcting the locations 
resulted in average location adjustments of 25-35 m. It was assumed that if the class 
occurred within these distances on the classified map then the site classification was 
correct. Accuracy assessment results are reported for each of these distance groups for 
the 1 1-, 13-, 17-, and 22-class maps in Tables 2-25 through 2-28. 






I" 



I * 



- S 

■ "5 1 



BB 



x 8 1 2- 
2 J - s 



2 J 4 * 



£ » 



08 GA S 



01 » 



■ 03 J» 



■*s 



c 
o 
•e 
a 
X 

■r 
V 

a 

U 



179 



r- 

o 



o 



o 



P 



a 8 

o a> 
00 .£ 



00 

K 



S 



o 



00 



o 



o 
V) 
m 



5 



00 



VI 



o 



vo 






00 



2 © 






•fi 

s 

i 

§ 
I 

X! 
E 



§• 



11 



00 

c 



a. 



s 



3 
O 

•S 
J 






a. o 




1 

E 
■a 

a. 
a 



■ 



.5 
a. 

o 
o 



.£> 

]5 
X) 
_o 

E" 

3 
&0 

J 

X> 

v> 
<S 
VI 

i 

.5 

Cl 
& 

A 



X) 
2 



fi 
e 
I 



Is 
II 

i) • 

la. 



E 


fr 


o? 


2 


VI 




1*. 


4> 


1 


T3 




i 



8 

B 
D. 

s 



13 
E 

« 

o~- c 

Al Q. 



U h 



1 - 

es — 

I 



« a 






s 

M — < 

< s 



a 
en 

Is 

2 g 

C c 



■ 

c 
£ 

■ 
■ 

a 

S 

S? 

© s? 



e 
r 
u 

E 

73 
o- 

J 

a. 



4> 

B 

£ 



- 
es 

i 
•a 
o 
e 



4 c 

B «« 

a. O 






U 



c s 



E 

a 
£ 65 



as 95 



• e -i 

9 H ^ ee C 

s 1 £ ■§ « 

« 9 5-5 

8 - I -o 



J £ « c. 

18 = 






re a 



c 



CU MB in 

C u — 



re c/, re — 



i j ■ 

08 ■ ■ -. 

_c! a* — — 



1 

•c 

i 



T3 

U 

1 



O 

o 

i 

i 
ro 
<N 

i 
<N 

<U 

I 



I 

u - 

1 i 

re = 

ex 

■ 



180 



so 
so 



i/-> 



s 

os 



s 

ON 
1/1 



SO 
SO 



M 

T 



SO 
SO 



S 

E 

> 

O 



O 
VI 

•6 



B 

E 



oo 
o 



s 

Al 



■t 



Xl 
in 

x 
_>> 

ir> 
<s 

VI 



t E 
2 2> 

s© 

r- .6 

Al Q. 



o 


!•• 


O 


rf 


m 


i/-i 


Os 


B5 


(S 


O 


Os 


o 


fN 


»— ' 


fl 



SO 

00 



VO 

oo 



<s 



AI.2 



OS 

o 



Os 
O 



i/-> 

Os 

Os 



Os 
OS 



. o 

I { 
11 

*s 

m X 

VI J 

1-8 
II 

1 & 

* 8 



E 



s 



oo 



Os 



in 



so 









r<-> 



to 



s 

1/1 






vs 

so 



ts 


W) 


o 


m 


JS 


O 


*m 


«-> 



<n 

SO 



1 

00 JS 

AIX 



£ 



sE 
VI 

•5 



o 



I 

X) 

x> 
x ■>, 

O C 

s i 

E § 



x 
2 

X 

<r\ 
<S 
VI 

8? 

X 



X 

o 



00 



R 



^J 



Alw 


os 

o 

00 


Al 







ex . 

O I 



5" 

as 

i 

a 
I 

I 

E 



a 
Q 

3 X 



X 

o 

-J _ 

I? 

7. GO 



x 

5 

X 
CA 

i 

»% 

re 

ea 






5 S^ 



i 

1 
ea 

■ 



5 

X 

j 

1 



I 



X 

e 

Si 
«s 



E 

z 

S 

£ 

U 2^ 



cl e 

Em cl 

m ~ I 






■ J -i 2 



U - 



5 £ * ft 
I 8 * 



J I 5 & 

• I if ^ 



« C 



K VI 



,5 » 



H fl . 

S_S CL 

2 T ^ 



Sin 
Bye, 



•O 
t> 

3 

c 
o 
o 

i 

(N 

JJ 



•i i 



181 



o 

o 



o 





p> 


f~ 


o 


00 


ve 


o 


00 


r- 


*-* 


cs 


m 




!•) 


IT-, 


»*> 



o 



00 



© 
o 

in 



o 



00 



o 



oo 



o 
o 

IT, 



© 



00 



u 

o 
-o 

I 

Cl 



I 



AI.O 



■6 



8 

J 



5 

Q. 

B 






o 


"a 




a> 


X> 




1 



■o 

e 
s 






13 •£ 
•a 



■ 


M 


L0 


■n 


in 


§ 


E 

C 


jf 


<*- 
o 


o 


3 


1 


1 


E 


T5 


1 


j 


Cl 

• 


a. 


S 


5 


t/> 



"ft -8 

U ■ — 

| J 

s i 



08 

■ 

a 

E 

>» 
Z 
£ 

OS 

j= 
Cl 



■ I ^N 

I M~ 

€ ® S 

g M 

■ E .a 

* s I 

-J < a. 









t» 



a 
O 



Cl O 



41 

B 

31 



c 



Cl 

en 



' - S 






182 



Table 2-24. Vegetation species found in ground-truthed sites used in the satellite image 
classification. 



Common Name 


Scientific Name 


Saw-Palmetto 


Serenoa repens 


Virginia Willow 


Itea virginica 


Blueberry 


Vaccineum spp. 


Oak 


Quercus spp. 


LongleafPine 


Pinus palustris 


Pond Pine 


Pinus serotina 


Slash Pine 


Pinus elliottii 


Loblolly Bay 


Gordonia lasianthus 


Sweet Bay 


Magnolia virginiana 


Blackgum 


Nyssa sylvatica v. biflora 


Pond Cypress 


Taxodium ascendens 


Red Maple 


Acer rub rum 


Titi 


Cyrilla racemiflora 


Fetterbush 


Leucothoe racemosa 


Hurrahbush 


Lyonia lucida 


Fragrant Water Lily 


Nymphaea odorata 


Spatterdock 


Nuphar luteum 


Bladderwort 


Utricularia spp. 


Dahoon Holly 


Ilex cassine 


Chain Fern 


Woodwardia virginiana 


Ogeechee Lime 


Nyssa ogeechee 


Walter's Greenbriar 


Smilax walteriana 


Bamboo Greenbriar 


Smilax laurifolia 


Walter's Sedge 


Carex walteri 



Table 2-24-continued. 



183 



Common Name 


Scientific Name 


Redroot 
Broomsedge 
Maidencane 

Gallberry 


Lacnanthes caroliniana 

Andropogon virginicus 

Panicum hemitomon 

Ilex glabra 



























184 




>. 






























w § 




























+■» 






























_c 


i 8 


^ 


8 


m 


«A 


S 


c 

£ 


g 


8 


8 


r- 


c> 






'5 


3 < 


vC 




O* 


00 












X 


r- 
































































EL 


1 1 

2 fi 


Cn 


f*i 


Tf 


<> 












>c 


«*i 






1 


S£ r- 


(N 


■<r 




C4 


— 


-* 


CN 


<N 


-~ 


■*r 


f> 
































£ 


>* 




























u 






























dtru 
ints. 


1 . •§ 
111 


r*> 


























§ 8. 


























































■ 




























2 § 

, s CO 




















O 




<N 




O &o 


*N 


















tt 




*T 






























£ 'a 






























_ o 






























o . 






























xels(10 
numbei 


SI 


















_ 


_ 




fS 






























•a s 


- - 




























^ <" 


cr 2 




























y 


< o 
















M 








rs 




C 3 
























































'£ eo 






























•c 


lis 

Li S. — 




























„ "o 




























§ U 














<N 










(N 




€ a 






























o eo 






























IB -O 


"3 ,. 




























assi 
ion 


Ell 

CQ 9 D 




























U CO 


























































» iS 


ft 




























eo 'Jo 


2 




























1 J 


3 
c 






























< j 




























ellite 
are c 










- 














- 
































e to 






























to 3 

to C 


c A 




























II 








fN 
















o 




tO 3 


CD co 


*r 






rs 












<N 


<N 


f*1 




|| 
























































V o 






























hell- 
data; 


$ s 




























CO JT 

> a 






f*1 


















*n 


























































•* <D 






























* g 

•5 u 


"9 

1 1 i 

o.?»c 








•*r 












en 


>n 


r- 




C (B 
























































CCS u- 






























Errorm 
ows are 


o2mu35 


Ov 






















o- 






























• « 


■ 


I! 

ill 


■ 










■ 


■ 

1 




-c 








Table 2-2 
location. 


u 
S 

•3 
3 
u 

■ 

> 


1 * 

■ T3 

11 
11 


v- 
U 

i 


i 

CO 

1 

'C 

m 


3 
II 


i 

cc S 


m 
| 8 


1 

o 

1 

< 


6 


1 

1 


CO 

CQ 

3 ■§ 
5 J 

-5 to 


H 

3 
U 



























Z a 

5 < 



a B 



1 ■ P 

9 5-2 

JEW 






185 






O so a. 



CO 5 






3. i 
< 3 



o 
o 

o 



11 

CQ w 



5 s 

*a 



> 

oo 



i I 



T3 
<L> 
3 
H 

C 

o 
o 

■ 
CM 

x> 

03 

H 



7 u 

- - u s 



3 S S a. 



o 2 & u m 

■ 
■ 

I 

I 

I 

> 



£ < 



SO 

00 



u 

2 

3 

o 
u 

< 
1 

> 

o 



186 



o 




S. 




o 








o. 




a 




CO 




<Si 




SZ 




+-» 




2 


c/i 


T3 


C 


C 
3 
O 


a 


O 


"a. 
E 


-~. 


CO 


£ 


V5 
Cm 


o 


O 


o 


k- 




<D 


s«^ 


X> 


(/2 


E 


<L> 


3 


X 


C 






Q. 




O 


CO 


PN 


GO 


c 


3 


'M 


CO 


£ 


> 






c 
o 









•m 




co 


CO 


o 


■*-* 


cG 


co 

T3 




c 


CO 


o 






o 


CO 


u 


o 


SP^ 


CO 




E 


C/l 




CO 



3 2 

U CO 



co 


GO 


GO 


c 


GO 


E 


eo 


9 


o 


o 


i 


CJ 


r«"> 




*-" 


« 


<U 


en 


J= 


-o 


1— 


CJ 


n 


Q 


Cm 


B 


X 


CJ 
u. 


V- 


« 


CO 


U 


E 


a; 




u. 


p 


eO 


h 


to 


w 


£ 




O 


»o 


OS 


<N 




(N 


c 
5 


3 


'■i 


CO 


| 






31 



-I 



II 



is 

3 



IJ 



j P s J 

u u. 7 J 



I 

Si 



6 S 

> 3 



Sill 

O m cj S 




so 

S3 

■c 

- 



<.3 



! II 

BOD 



cj go 



£<§ 














187 




| 














J - ™ 














ll 


*c 












3 < 


ac 










ll 


■*) 








a w 


f> 










X 










*] 














o tw 














31 






8 








K -° 














11 




*S 


iTl 








U m 




f 


O 






*J 




(N 


o 


tJ 






I 




< a 




IN 










hi 






8 








U in £ 




<N 




1 




ill 






8 




■ _ 




— 


"" " 








|j 






8 






11 










si 




O 


f*t 








(N 


f*1 


r» 


00 






















o 






&ii & 








o 












o 






ni* 




2 


o 

00 


© 
II 




















1 








u 






II 




fS 


VO 


> 






tn 


n-> 


VI 
















00 














II 

S2 




f*1 


IB 


o 

II 


















h 








df 
E 










vi 


o 






_ — 




rj 


r^ 


oo 




3 












II 




C 














mm 




1 








>> 




C 
O 

o 




liifl 




o\ 


8 


1 

3 

8 




vo 














i 






j$ 




■ 


< 




ts 




s 


c wo 

■8 fr 

J CD 




V U 


S3 




m 

N 




1 

> O 


ll 

5(2 


g | 
I 1 


■ 
E 

> 

o 

















188 



g 

03 

M 

■a u 

■a 3 

8.1 

— s 

Cl «? 

E^ 

CO 

■g | 

§ « 

o ja 
E iu 

O e3 
O OT 

X o 
'5- | 

O «3 

•* 03 
C T3 
O 
O 

c 

& 

| 
s 

« •- 

03 <U 

° !3 

•55 ^ 
22 2 

03 o 

u 

00 i 



c 

o 



s 



m 
•E cs 
j| s 

= 03 

<U t— 

■*-• ' 

03 o 

.2 eg 

4) Xi 

•S E 

c§ C 

1 " -S 

E c c 

t u E- 

pq do c 

U 03 

1> ** (M 
<N O 

cn c h 

°* o a> 

« • -3 X) 

3 S E 



§ 1 

D < 


ON 


§ 


8 


O 




00 


Q 
00 


m 


8 


8 


8 


8 


Q 

O 




00 






* 3 

M H 


o 




(N 


■"t 






© 


l> 






<N 


fS 




O 


IT) 


c*i 


OS 


<N 
C4 










*f 






— 












— 


» 


— 


v^ 


(S 










HM 


















fN 




o 




On 

oo" 










rs 




































*N 


















(N 




(N 




*r 


























_ 




_- 






*^i 
























IN 












<N 






















fN 














- 




















. 
















O 


















_ 


















o> 
















* 














- 




fS 


00 








CN 






00 


































00 
















u*> 


f*> 




(N 


VO 










*s 


























vn 








(N 






~ 






















f*> 






<N 




in 


























(S 




„ 
































* 


o 










in 




, 




















Vegetation 

Class 

Number 


^r 


rs 


f> 


tfl 


<o 


in 


00 


On 


o 


mm 


rsi 


T*> 


*r 




o> 
oo" 


O 
<N 





189 



u 

3 

a 

c 
o 



t 

<N 

4> 
I 



5^ 

D < 






ll 

m H 












O 


f^ 




oo 


^ 






O 


VI 


■V 


(S 


Q 

v-i 


f«"t 


rj 


8 


fS 


rs 


S 


- 


, 


8 


o 


. 


8 


c* 


h» 


in 


00 


c 


o 

00 






^0 

VI 


^c 


rN 


8 


</i 


*•*, 


is 


W 


P 1 " 




rs 


„ 


8 


^r 




£ 

r^ 


Vegetation 

Class 

Number 


l 1 

Q h 


■ 
V to 

i s 



o 
o 



> 



^ 



r>- 



3 
U 

< 

u. 



a 

.2 ^ 

Se 

O I- 

.2 Jg 

1 I 

•a 3 

& u 

u a 

•JB « 

£ y 

M 73 

5 > 

g | 

O o 



o !£j 

1— c/i 
» — ' on 

73 " 

X u 

■j; S 

c o 

2 £ 

c g 

II 

c $ 
.2 1 

S3 <L> 

O 

ss g 

</) c 
CO CfiS 

o e 

150 a 

CO « 

•9 * 

<D o 

i * 

1 «? 

I » 

*§ 

« o 

<u *-» 
_C on 

♦-> u. 



190 



1 i 

a c 
oo e *a 

^.2 a 

JJ f 73. 

7a «> S 
H > I 



[A * 


s 


1 


© 
5 




o 


5 


00 


o 
oo 




s 


s 


g 


8 


O 

o 


o 


1 


o 










r< 


„ 


r* 


00 


*t 




CN 


o 


p* 


„ 


_ 


rs 


M 


M 


r- 


r* 


00 


9* 


10 


(*i 














-* 






- 
















- 


- 




M 


r» 














PM 




m 
















M 








O 












m 






















C* 






o 


O 












m 
























r* 


3 




OO 












- 


























^ 




r- 


































o 






p. 


£ 












r* 




















(N 


- 






- 


«n 








- 






(N 
























p. 




^ i 




























~ 








- 






ci 


























r* 
















<s 
























<M 


















- 






















- 




















o 




















- 






















Ov 


















•*t 


















- 






00 










M 






00 


























r^ 












M 


r- 


PM 














Tt 




f> 


c« 






10 












c* 






























Ti 










<s 






_ 


























T 








o 






is 




- 












- 












**) 






*s 






T> 






























C* 




- 






































- 


fS 






_ 






















- 












Vegetation 

Class 

Number 




r< 


pi 


n 


m 


« 


r^ 


00 


OS 


o 




r* 


m 


■*■ 


Wl 


10 


r- 


00 


<h 


o 















191 




sl 


wi 


Os 












2 < 


r- 


r- 






















11 




os 












OS H 


^ > 












« 




ws 


ri 


«ft 




n 




*■* 


r< 


« 














tn 




(S 


#»i 




r- 


■<r 






o 








r* 




C* 






"" 


(-- 






OS 






*r 


Os 










— ■ 


1^ 






00 






r- 






"" 






fS 


00 






r- 






© 


B 


sO 








r*s 










<■ 


f»> 






>rt 












"" 




■■ 


<n 


o 






^f 








o 




~ 






CM 


»n 






ri 








S 




" 






CN 


~ 






f* 








8 










CS 
















8 










*~ 


~ 






o 








8 










"" 


— 














i- 




Os 




r* 


r~ 


«r» 


















© 












o 


© 


o 






00 






- 


00 


o 
II 






















o 


«n 


S~\ 






I- 




m 


(N 


m 


u 






















(N 


8 


> 






so 






— 




so 
















v% 






f*l 


SO 


© 
II 


















^ , 






© 


1 


u 
















g 


















n 






t^ 




so 




3 














II 




C 




r* 








8 




B 

o 














i 










*n 


O 


00 

I 






" 




" 


00 


o 


< 






















tation 
ber 






. 


II 


E 




s 




M 1 






|J 


*6 3 


6 




H 




> O 2 


fS 


cm 


i% 


£ 



















192 
Overall map accuracy was computed by dividing the number of correctly 
classified sample sites by the total number of sites. User's and producer's accuracies 
were calculated for each matrix. User's accuracy (errors of commission; # correctly 
classified sites in a category / total # sites placed in that category) indicates the likelihood 
that a classified pixel actually is that class on the ground. Producer's accuracy (errors of 
omission; # correctly classified sites in a category / # reference sites in that category) 
indicates whether reference sites were correctly classified (Lillesand and Kiefer 1994). 

Reliability of the classification can be assessed with the Kappa coefficient of 
agreement (K). The coefficient measures whether agreement between the actual points 
and the classification is true or due to chance (Lillesand and Kiefer 1994), and aids in 
determination of sources of classification errors (Fung and LeDrew 1988). The K for 
each map and classes with >18 sample sites were calculated using spreadsheet software 
and equations documented by Hudson and Ramm (1987), Rosenfield and Fitzpatrick- 
Lins (1986), Congalton et al. (1983), Rosenfield et al. (1982), and Cohen (1960). The 
swamp upland islands were not included in these points. Pair-wise comparisons 
between Ks of various classifications were calculated to determine significantly different 
error matrices, using the formula 

z^-K.yrvtK^v^)]* 

where K is the kappa coefficient, V(K) is the coefficient variance, and Z is the standard 
normal deviate. Significantly different K's (at the 95% confidence level) are indicated 
where Z> 1.96. 



193 
Agreement between the reference data and the classified map was also calculated 
for individual classes as a measure of categorical accuracy using the formula 

K c =(NX li -X i+ X +1 )/(NX 1+ -X, 4 X +I ) 
where K,. is the accuracy measure for a given category, N is the total number of counts, 
Xjj represents matrix cell totals, and X I+ and X+j represent row and column totals, 
respectively (Rosenfield and Fitzpatrick-Lins 1986). The categorical K was calculated 
only for those classes with a minimum sample size of 19. At this sample size, the 
probability that the map was correctly classified for a given class is 85% (Rosenfield et 
al. 1982). 
Image Classification and Accuracy Results 

Class composition and class areas for each map (1 1-, 13-, 17-, and 22-class maps) 
are detailed in Table 2-23. A list of vegetation species recognized in the classified image 
is given in Table 2-24. Blackgum (Nyssa sylvatica v. biflora), cypress {Taxodium 
ascendens), and loblolly bay (Gordonia lasianthus) classes cover approximately 75% of 
the swamp. Error matrices for the 1 1-, 13-, 17-, and 22-class maps are found in Tables 2- 
25, 2-26, 2-27, and 2-28. The overall accuracies are 86%, 81%, 75%, and 67% for the 
1 1-, 13-, 17-, and 22-class maps, respectively. Comparisons of the Ks for these maps 
showed that the 17-class map was not a significant (P=0.56) improvement in 
classification accuracy from the 22-class map (K 22 =0.65, K 17 =0.72); the 1 1- and 13-class 
maps (K n =0.82, K 13 =0.78) were significantly more accurate than the 22-class map 
(P u =.0002, P 13 =.0018), although the improvement was not significant when 



194 
consolidating from 17 to 13 (P=0.62) or 13 to 1 1 classes (P=0.22). Categorical Ks are 
given in Table 2-29. The blackgum, cypress, loblolly bay, wetland pine (Pinus serotina, 
P. elliottii), and shrub classes [titi (Cyrilla racemiflora), hurrahbush (Lyonia lucida), 
fetterbush {Leucothoe racemosa), Virginia willow (Ilea virginica), dahoon holly (Ilex 
cassine), bamboo briar (Smilax laurifolia), Walter's briar (Smilax waiter •/'), wax myrtle 
{Myrica cerifera), soapbush (Clethra alterniflora)], which constitute approximately 90% 
of the classified image area, had sample sizes sufficient for categorical calculations (KJ. 
Greatest categorical improvement occurred when consolidations were within blackgum, 
wetland pine, and loblolly bay classes, which cover 52% of the classified map. Pine 
(wetland and upland) and shrub classes were the most frequently misclassified 
categories, most often confused with blackgum and cypress classes (Table 2-30). 

Although the overall accuracy for the 22-class map is only 67%, there are 1 5 
classes with user's accuracies >75%, which means that the probability that a classified 
site is truly that class is >75%. Problematic classes make up 47% of the classified area, 
predominately mature bay-shrub, gum-bay-cypress-shrub, and scrub cypress-shrub. 
User's accuracies for those classes ranged from (?ine-Woodwardia) to 69% (scrub- 
cypress-shrubs). In the 17-class map, with overall accuracy of 75%, there are 13 classes 
with user's accuracies >75%; 38% of the classified area contains classes that have lower 
accuracies (Nuphar-Nymphaea, gum-bay-cypress-shrub, S>w7ax-shrub, and bay-shrub). 
Improvement to >75% user's accuracy for these classes does not occur until they are 
consolidated to 1 1 classes. In the 1 1 -class map, only gum-maple-bays (K c =66%) has 
user's accuracy <75%. 



195 



Table 2-29. Categorical Kappa coefficients (K c ), user's and producer's accuracies for 
classes with >18 ground-truthed sites in the 11-, 13-, 17-, and 22-class classifications 
within 10 pixels (100 m) of sample location. 



Map, Vegetation Class 
Name and Number 


K,(%) 


User's 
Accuracy (%) 


Producer's 
Accuracy (%) 


11-Class Map 








Gum-Maple-Bays- 
Cypress-Shrub (3, 6) 


61.9 


66 


100 


Upland and Wetland 


100.0 


100 


75 


Pines (1, 4, 7, 15, 21) 








Briar-Shrub (9, 22) 


81.9 


85 


73 


Cypress-Shrub (2, 18, 
19) 


83.4 


87 


95 


Loblolly Bay-Shrub (16, 

17, 20) 


75.1 


79 


90 


13-Oass Map 








Gum-Maple-Bays- 
Cypress-Shrub (3, 6) 


61.9 


66 


100 


Upland Pine (1,4) 


94.3 


95 


76 


Wetland Pine (7, 15, 21) 


74.1 


78 


56 


Briar-Shrub (9, 22) 


81.9 


85 


73 


Cypress-Shrub (2, 18, 
19) 


83.4 


86 


95 


Loblolly Bay-Shrub (16, 
17,20) 


75.1 


79 


90 


17-Class Map 








Upland Pine (1,4) 


94.3 


95 


76 


Wetland Pine (7, 15, 21) 


74.1 


78 


56 


Cypress-Gum-Shrub (18, 
19) 


83.2 


87 


95 


Shrubs (22) 


76.2 


79 


65 


Gum-Bay-Cypress- 
Shrub (6) 


40.9 


44 


100 


Bay-Shrub (16, 17) 


56.5 


60 


75 



Table 2-29-continued. 



196 



Map, Vegetation Class 
Name and Number 


«,(%) 


User's 
Accuracy (%) 


Producer's 
Accuracy (%) 


22-Class Map 








Gum-Bay-Cypress- 
Shrub (6) 


40.9 


44 


100 


Mature Cypress-Shrub 
(18) 


72.1 


76 


81 


Shrub (22) 


76.2 


79 


65 






197 

Table 2-30. Class confusions in the 1 1-, 13-, 17-, and 22-class classifications within 10 
pixels (100 m) of sample location, for classes with user's accuracy <80%. The class with 
most frequent error is underlined. 



Map, Vegetation Class 
Name and Number 


User's Accuracy 

(%) 


Class Confusions Occurring in Map 


11-Class Map 






Gum-Maple-Bays- 
Cypress-Shrub (3, 6) 


66 


Pines. Shrubs. Cvpress-Shrub. 
Loblolly Bay-Shrub 


Loblolly Bay-Shrub (16, 

17, 20) 


79 


Pines. Shrubs 


13-Class Map 






Gum-Maple-Bays- 
Cypress-Shrub (3, 6) 


66 


Wetland Pine. Shrubs. Loblollv Bav- 
Shrub, Cypress-Shrub 


Water Lilies (5) 


50 


Carex-Nymphaea 


Wetland Pine (7, 15, 21) 


78 


Upland Pine 


Loblolly Bay-Shrub (16, 
17, 20) 


79 


Wetland Pine. Shrubs 


17-ClassMap 






Nuphar-Nymphaea (5) 


50 


Carex-Nymphaea 


Gum-Bay-Cypress- 
Shrub (6) 


44 


Gum-Maple-Bavs. Wetland Pine. Bav- 
Shrub, Cypress-Gum-Shrub, Mature 
Loblolly Bay, Shrub 


Wetland Pine (7, 15, 21) 


78 


Upland Pine 


Briar-Shrub (9) 


57 


Upland Pine, Wetland Pine, Shrubs 


Loblolly Bay-Shrub (16, 
17) 


60 


Wetland Pine. Mature Loblollv Bay 
Shrub 


Loblolly Bay (20) 


77 


Bav-Shrub. Shrub 


Shrub (22) 


79 


Wetland Pine, Briar-Shrub 


22-ClassMap 






Pine-Palmetto (4) 


75 


100% Upland Pine, Pine-Gum-Bay 


Nuphar-Nymphaea (5) 


50 


Carex-Nymphaea 



Table 2-30--continued. 



198 



Map, Vegetation Class 
Name and Number 


User's Accuracy 
(%) 


Class Confusions Occurring in Map 


Gum-Bay-Cypress- 
Shrub (6) 


44 


Gum-Maple-Bavs. Pine-Gum-Bav, 
Young Bay-Shrub, Mature Cypress- 
Shrub, Scrub Cypress-Shrub, Mature 
Loblollv Bav. Shrub 


Pine-Gum-Bay (7) 


58 


Pine-Palmetto. Pine- Woodwardia. 
Wetland Pine 


Briar-Shrub (9) 


57 


Pine-Palmetto, Wetland Pine, Shrub 


Pine Woodwardia (15) 





100% Upland Pine, Pine-Palmetto, 
Pine-Gum-Bav 


Mature Bay-Shrub (17) 


50 


Pine-Gum-Bav. Young Bav-Shrub. 
Mature Loblolly Bay, Wetland Pine, 
Shrub 


Mature Cypress-Shrub 
(18) 


76 


Pine-Gum-Bav. Smilax-Shrub. Open 
Water. Scrub Cvpress-Shrub. Shrub 


Scrub Cypress-Shrub 
(19) 


69 


Pine-Woodwardia. Mature Cvpress- 
Shrub 


Mature Loblolly Bay 
(20) 


77 


Young Bay-Shrub, Mature Bay-Shrub, 
Shrub 


Wetland Pine (21) 


75 


100% Upland Pine 


Shrubs (22) 


79 


Pine-Gum-Bav. Briar-Shrub. Pine- 
Woodwardia 



199 
Reference sites with producer's accuracy >75% in the 22-class map (12 classes) 
included those comprising approximately 80% of the swamp. This indicates that the 
reference pixels for the classes covering a greater proportion of the swamp were fairly 
accurate examples of those vegetation types. The 10 classes with lower producer's 
accuracy are pine, loblolly bay, and shrub classes that are consolidated in the 17-, 13-, 
and 1 1 -class maps. Only the open water and shrub classes, with producer's accuracies of 
50% and 73%, respectively, remain <75% accurate in the 1 1 -class map. This is probably 
because of location error; one of the open water sites was in a cypress pond and was 
misclassified as the surrounding cypress class. The shrub class often occurs in the matrix 
among blackgum, cypress, and loblolly bay classes, where an error of 50 m could result 
in a perceived location in a different vegetation class. 

Although island ground-truth points were not included in the original 
classification, it was recognized that errors probably existed in these areas. The swamp 
upland island area classification was improved with the subset and reclassification. 
Accuracies in the reclassification area are reported in Table 2-3 1 . 
Interpreting the Accuracy Assessment 

User's and producer's accuracy statistics can be used in map interpretation to 
identify classes with greater error likelihood. Map interpretation is aided by knowing the 
causes of classification errors. Misclassifications may be the result of location error 
(difference between ground-truthed location and map location), observer bias when 
estimating proportions, or man-induced changes in the site between the image capture 



200 



Table 2-31. Error matrix of classes in the swamp-and-island-uplands map that are 
combined with those in the 17-class map. Cell values are number of samples re- 
classified in the swamp-and-upland-islands classification from classes in the 17-class 
map. 









Mixed 








Vegetation 
Class 


Pine-Cypress- 
Hardwoods 


Upland and 

Wetland 

Shrubs 


Sparse Pine 


Dense Pine 


User's 
Accuracy 


17-ClassMap 














Bay-Shrub 


1 




1 


4 


4 


10 


Upland Pine 








6 




100 


Wetland Pine 








6 


4 


100 


Shrubs 








2 


1 





Sedges-Ferns- 
Water Lilies 


2 






2 







Briar-Shrub 






2 






100 


Cypress-Gum- 
Shrub 


3 






2 


2 


43 


Gum-Bay- 
Cypress-Shrub 


1 




1 






50 


Loblolly Bay 












100 


Gum-Maple- 
Bays 








1 







Producer's 
Accuracy (%) 




75 


100 


52 


57 





Overall accuracy of additional classes in swamp and island uplands map= 75%. 



201 
and ground-truthing dates. Identifying these types of errors requires recognition of the 
likely successional sequence, awareness of typical land use practices that might change 
the landscape composition, such as clear cutting and re-planting timber, and familiarity 
with the region and vegetation being classified. 

Errors that occurred in the 1990 swamp image classification usually involved 
pine, shrub, and mixed blackgum classes. Most classes on the image contain some 
proportion of pine and shrub; if the location was in error or if the patch was not evenly 
dense and this was detectable at the 10 m pixel level, then the pine and shrub classes may 
have been recorded. The mixed blackgum classes are combinations of many species. 
When blackgum classes are highly interspersed with other classes, and location error 
occurs, the classes may be misidentified. Errors in the upland island classification also 
involved confusions of pine and shrub classes. Island classes were delineated and 
identified by refuge foresters. Although they would like to distinguish among pine 
densities and interspersion with hardwoods, there may have been insufficient differences 
in spectral signatures of these types because of species interspersions in their selected 
reference sites. Delineation of upland pine and shrub community types may be possible 
only between "pine" (where pine-cypress-hardwoods and dense and sparse pine classes 
are combined) and "not pine" (represented here by grasses and shrubs). 

The most accurate maps were the 1 1- and 13-class maps, which were not 
significantly different. Accuracy of the 1 1- and 13-class maps were significantly better 
when searching a radius of 100 m than 35-50 m for specific class types. The number of 
correctly classified sites on the 22- and 17-class maps did not depend on this search 



202 
radius. Map complexity or patch interspersion may affect perceived map accuracy; the 
map may actually be correct, but location error misplacing an observer on the ground 
suggests map error. Reporting map accuracy by distances of search radii around the 
target pixel(s) incorporates information about the location error of ground-truthed sites, 
and aids in identifying if the error is a true classification error, the result of class 
patchiness, or due to location error. 

A pplying Image Classification Procedures 

Using a classification scheme based on spectral rather than textural qualities also 
aided in the classification accuracy. The spatial complexity of the swamp vegetation 
requires this fine data resolution; swamp and upland island vegetation communities 
occur in a mosaic rather than in large, single-species patches. Detection of the 
interspersed vegetation types is compromised at the 30 m pixel level. Even with 10 m 
pixels data are lost; details at sub-pixel level are not detectable, which must be 
considered when assessing vegetation community changes over time. This spatial 
complexity also affects map accuracy. Although the map accuracy assessment can be 
automated, the ground-truthed sites should be examined on the image to identify the 
types of errors occurring e.g., location error, class confusion, class overlap. 

A thorough understanding of the classification process is necessary to use the 
classified map properly. For this image classification there were several upland classes 
that did not occur in the wetland part of the swamp. These classes were eliminated from 
the training set used to classify the swamp proper, which was extracted from the image 



203 
and then stitched back to the perimeter area after classifying. Selecting a reduced 

signature set was also necessary in the matching region, where the north and south 

images were joined. The classification accuracy would have been improved initially by 

selecting training sites isolated on the large islands. These islands contain upland shrub 

species [saw-palmetto {Serenoa repens), gallberry {Ilex glabra), blueberry (Vaccineum 

spp.), oak (Quercus spp.)] not found in the wetland environment; because they were not 

included in the original training signature set, the classified map shows shrub 

communities on the islands, but their species composition differs from those in the 

swamp. Other classification errors such as confusing classes, could be remedied in an 

iterative process of classifying, ground-truthing, re-classifying, ground-truthing, etc., 

which was cost-prohibitive in this study. Change assessments must recognize if this type 

of classification signature set manipulation has occurred, so that changing class 

composition and distribution are recognized only where they occurred naturally. 

Selection of representative signatures in developing the training set for supervised 

classification requires a thorough knowledge of the area of interest. Since the 

classification will be forced to assign pixels to the specified class selection, the 

signatures must be typical of all classes present. This is facilitated by beginning with an 

unsupervised training site selection and image classification, and using the results to 

locate class types and locations for ground-truth identification. The process can then be 

repeated using those ground-truthed sites in supervised classification seed set selection 

and the supervised classification of the image. Subsequent ground-truthing will provide 

data for an accuracy assessment. Time and budget limitations prevented ground-truthing 



204 
more than one set of sites. Rutchey and Vilcheck (1994) found that the spatial 
heterogeneity of the Everglades, which is probably similar to that of Okefenokee Swamp, 
limited the number of classes in their unsupervised classification; they used a minimum 
class patch size of a 3x3 pixel (60 m X 60 m) window to eliminate classes that occurred 
only in smaller patches. Because some categories did not occur in patches greater than 
60 m X 60 m, they overlooked some classes. They suggest beginning with a large 
number of classes, without the patch size restriction, and combining them until the 
desired accuracy is achieved. This is essentially the procedure followed in this 
supervised Okefenokee classification. 

Class consolidation involves combining mixed and/or single species classes. In 
this classification groups with similar species but different proportions were combined 
when the maps were consolidated. Accuracy improvements were not significant for the 
consolidation from the 22-class map to the 17-class map, but improvements were 
significant for the 17- to 13- and 17- to 1 1 -class combinations. The groupings from 22 to 
17 classes concerned some mixed species classes, but left some classes with the same 
dominants separate. The accuracies of these classes improved when lumped in the 13- 
class map. These groupings involved mixed species and ages with the same dominant 
but different sub-dominant species (e.g., young loblolly bay-shrub combined with mature 
loblolly bay-shrub and mature loblolly bay). Upland island classification accuracy would 
also have improved by combining the pine-dominated classes into one class. Even the 
more homogeneous classes are complex enough to cause some confusion with mixed 



205 
classes in a 10 m pixel image due to location error. The highly interspersed matrix 
requires a small location error to correctly assess classification accuracy. 

The limitations of imagery to detect sub-pixel changes must be recognized in 
change detection studies. Aerial photograph resolution may permit identification of 
more detail than from SPOT 10 m resolution imagery. Interpretation of aerial 
photography includes spectral as well as textural information, which can aid in class 
identification. Definition of edges in satellite imagery is difficult where mixed class 
composition varies. Stereo color infrared aerial photographs may be better tools than 
imagery for mapping the mixed classes, since species details may be discernable at the 
sub-pixel (100 m 2 ) scale. However, spatial registration of photographs may be difficult 
due to photograph distortion and absence of reference features in wilderness areas (see 
Chapter 4). Imagery is more easily geo-referenced due to its greater spatial extent and 
likelihood of including landscape features suitable for registration. Additionally, satellite 
image data provide a large amount of information which can be rapidly processed using 
computer discriminated vegetation types. Detection of changes in community 
composition and distribution might be most accurate when aerial photographs and 
satellite images are interpreted together (Silveira 1996). 

The maps produced here are sufficiently accurate for change detection study 
within the swamp if the following are recognized: 

1) The classification procedure selectively included classes in the signature 
set; absence of a class in a particular area could be due to its exclusion from the signature 
set, although signature eliminations were done primarily for highly unlikely classes (e.g., 



206 
upland classes removed from classification of wetland areas and wetland classes 
removed from upland areas). 

2) Classes are not equally accurate; shrub, pine, and blackgum classes may 
be confused with other mixed species classes, and classes based on species' densities 
may be misidentified. 

3) Detectable changes with satellite imagery will be limited to the scale of 
10s of meters given the image pixel size (10 m). Aerial photography should be used to 
detect changes at the sub-pixel level. 



CHAPTER 3 
OKEFENOKEE SWAMP HYDROLOGY MODEL 



Introduction 

The Okefenokee Swamp hydrologic environment has a history of manipulation. 
Although indirect impacts to the hydrology were occurring as settlements arose in the 
surrounding landscape and wildfire control, prescribed burning, grazing by domestic 
stock, and timber harvest were increasingly practiced during the 18 th and 19 th centuries, it 
was not until 1890 that the direct assault began (Trowell 1989c). Attempts to drain the 
swamp failed, but the excavation left a 20 km ditch (the Suwannee Canal) connecting the 
eastern shrub and prairie environments to the western river system. The extensive 
logging following the drainage attempt removed timber from 26% of the landscape (see 
Chapter 4), and the composition and structure of vegetation in the landscape changed 
with vegetation regrowth (see Chapter 4). Inhabitation of the surroundings increased the 
perceived need to control wildfire, which was a vital process in the dynamics of the 
swamp and the perimeter upland vegetation communities. In response to personal 
property and perceived ecological damage caused by widespread fires in 1954-1955, the 
Suwannee River Sill was constructed in 1960 to impound the swamp and protect it from 
future drought and fire (Chapter 742, Public Law 81-810, 70 Statute 668). The decades 

207 



208 
that followed did not show a decrease in fire frequency (see Chapter 5), and the 
realization by swamp managers and ecologists that wildfire was integral to the system's 
health challenged the declared purpose of the Suwannee River sill. In 1990 review of the 
sill's purpose and actual effect on the swamp hydrologic environment and vegetation 
communities was determined necessary, before repairs or changes to the decaying 
structure could be recommended (Roelle and Hamilton 1990). A spatial computer model 
would provide temporal and spatial information about the Sill's area and degree of effect 
on swamp hydrology, and permit manipulation of the swamp landscape and hydrologic 
features to identify the system's sensitivities. It was for those purposes that the 
hydrology model discussed herein was developed. 

This chapter discusses development of the spatial model of the Okefenokee 
Swamp hydrologic environment (HYDRO-MODEL), manipulations of model parameters 
that suggest system sensitivities, indications of the Sill's impact area, and extent of 
effects of the existing Sill identified with model manipulations. Model application and 
analyses focus on the following questions: 1 ) Has the sill changed the swamp hydrologic 
environment? If so, where and how have these changes occurred? 2) Have vegetation 
changes reflective of the sill's influence occurred disproportionately in the area affected 
by the sill? 3) Have wildfire size and frequency changed in the area impacted by the sill? 
4) What changes in swamp hydrology and vegetation distributions can be anticipated 
with the sill's removal? Model code and detailed instructions for implementation are 
included in Appendix B. Development of model databases are briefly discussed in this 
chapter, and in detail in Chapter 2. 



209 
Methods 

Model Objective 

The Okefenokee Swamp hydrology model was developed with weather and 
vegetation data representing conditions in the swamp during 1980-1993, and topographic 
information collected during 1991-1994. The model is intended to represent the swamp 
hydrology cycling in twice-monthly time steps during 1980-1993, and provides output in 
sample point form and water surface elevation and depth maps of the Okefenokee 
National Wildlife Refuge area for each process interval. Output data include water 
depth, water surface elevation, and amount of water moved in each time step, and can be 
viewed by individual interval, in a "movie" of the entire process period by monthly 
intervals, and queried as entire maps or individual cell values. The model was built and 
calibrated using data from 1980-1993, and run with independent data sets for decade 
intervals of 1941-1949, 1950-1959, 1960-1969, and 1970-1979 to assess model 
performance. The "with-" and "pre-sill" conditions (1960-1993 and 1941-1959, 
respectively) were represented by topographic surfaces with and without the sill in place. 
Data from 1980-1993 were also applied in the model to the no-sill topographic surface to 
demonstrate water surface elevations that might have occurred during that period had the 
sill been absent. The model is a predictive tool in that input data grids can be modified 
to reflect potential changed conditions (such as no sill, no ET, no precipitation) and 
output grids compared, and it should be used to examine trends in water surface 






210 
elevations over time. The model can not predict future conditions since it relies on 
actual, recent flow, evapotranspiration, and precipitation data summarized semi-monthly. 
Examination and comparison of current weather and swamp hydrology, and trends and 
conditions during previous months, seasons, or years should provide an indication of the 
potential swamp hydrologic environment that could be expected during any month with 
various weather conditions. Its use in this study was not to predict the current or future 
swamp hydrologic environment, but to identify the region of the sill's impoundment 
effects and how the recent hydrologic environment of the effected area might have 
differed in the sill's absence. 
Model Overview 

HYDRO-MODEL is written in ARCINFO Macro Language (AML) (version 7.0, 
ESRI, Inc., Redlands, CA 92373) routines to operate in the ARCGRID Unix 
environment. The complete model text is provided in Appendix B. The model is a grid- 
cell model that processes within the Okefenokee National Wildlife Refuge boundaries. 
Each cell in the landscape encompasses 250,000 m 2 (500 m x 500 m); 10,672 cells are 
modeled (Figure 3-1). After the model initiates processing by setting user-defined 
parameters and interval dates, several processes occur within each cell (Figure 3-2) 
during 3 main model phases. In Phase I a water surface is created (inh20xcr, where xxx 
specifies the year, month, and interval for processing) by combining a starting water 
depth that is defined either by the decade starting date or created in the final processes of 
the previous interval's Phase HI, the swamp topographic surface elevation, and inflowing 



211 



TO 
■D 

C 



W 



8r8§ 



m 





CD 



2 



w 


o 

i 

o 



a 
| 

on 

| 

o 

c 

O 



u 
t- 
ed 

00 

c 

• •■* 

I 

I 



I 



I 



212 



E 
cq 

■a 
o 

£ 
I 




1 

§ 

a 
I 

u 

W 
Q 
O 

i 

o 



a 
E 

on 
a> 

o 
c 
•fi 

O 



i 

| 

o 

u, 

i 

m 
g 



213 
water from perimeter creeks. Creek or river outflow are removed from appropriate cells 
and evapotranspiration occurs in each cell, creating a net water surface (netxxx, where 
xxx specifies the year, month, and interval for processing) in Phase II (Figure 3-2). 
Surface sheetflow occurs in Phase III. If the water surface is sloped, determined by the 
topographic surface elevation, water depth, and a query of neighboring cells (Figure 3-3), 
water is moved to the neighboring cell with the lowest elevation. The amount of water 
moved in this step is determined by a subroutine that identifies how much water should 
move in and out of each cell based on local elevation gradients. If a gradient is flat, no 
water movement occurs in that cell's immediate neighborhood. The model processes 
this sequence for a user-defined number of iterations ("# of pixels to move water" in the 
model interface, Figure 3-4) . This permits water to move more than one cell length in a 
semi-monthly interval, or to move at single cell lengths if data are provided for shorter 
model iterations. The final model products of Phase III are ending water surface 
elevation (ewatxxx) and water depth (dwatxxc, where jcxx specifies the year, month, and 
interval for processing) for the specified interval. The water depth surface (dwatxxx) 
then becomes a starting surface for Phase I of the next model iteration. 

Several modifying grids were added to the basic processing steps listed above to 
refine water movement in the swamp landscape. Inflow into the swamp occurs primarily 
in the Northwest from streams that flow continuously but with seasonal fluctuations. 
Water from these streams most likely flows along the topographic gradient on the 
western swamp to the Suwannee River and does not cross over to the eastern swamp 
because of topographic slope. Therefore, to direct movement of this water that originates 



214 



Start at cell address 0,0: 

1 ) Check flow movement direction 
at each neighboring cell address. 

2) Sum water moving into from 

each neighbor. 

3) Move to cell 1 ,0. 

4) Repeat across entire grid for 
"pixels-to-move-water" iterations. 

5) Report "amt-to-move". 

6) Return to Phase 3. 



Cell Address 



Movement 
Direction 



-1,1 


0,1 


1,1 


0,-1 


0,0 


1,0 


-1,-1 


0,-1 


1,-1 



128 



64 



32 



16 



T 



8 









Figure 3-3. Neighborhood search in HYDRO-MODEL to determine direction and 
amount of water to move in each cell and time interval. 



215 



Okefenokee Hydro Model 




HYDRO-MODEL 



Start Year 



1941 1 1950 1 1960 1 1970 1 1980 1 1990 



Ending: 



Month 

12 



Year 

1993 



Percent PET to Use 

April-May, Oct-Nov Little Rain, High Evap: 




1 0.000 

June-Sept, High Evap, High Rain 
1.25 0.000 

Dec-Mar, Average Rain, Little Evap 
1.50 0.000 



J 1.50 





I 1.50 



1.50 



Percent Water to Move 



Inflow Zones 
.01 0.000 




J 0.020 



Suwannee Outflow Adjustment 
.20 



# of Pixels to Move Water 
Pixel Size is 500 Meters 

10 



(Apply ) (Cancel) 



Figure 3-4. HYDRO-MODEL menu interface for setting user-defined parameters. 



216 
in the Northwest creeks and flows through the western swamp and river floodplain, zones 

were created (Figure 3-5); in each model iteration the estimated flow into each zone, 
representing inflow by creek drainages, is proportioned equally among the zone's cells. 
This permits movement of the entire inflowing volume through the landscape with each 
model iteration. 

Outflow is similarly proportioned in an outflow zone near the Suwannee River 
Sill (Figure 3-6), and removed in each interval. Outflow zones for other exiting flows 
(St. Marys River, Cypress Creek, and Sweetwater Creek) are also coded into the model 
so that flow volumes can be incrementally removed; however, topographic gradient was 
used to move water in these areas in the model iterations discussed here, and the creek 
outflow zone code was bypassed. Water also flows over the sill, directed by the surface 
gradient. The estimated flow volume in the inflow and outflow zones can be adjusted by 
proportional multipliers (for inflow, "Percent Water to Move", and for outflow, 
"Suwannee Outflow Adjustment"), to fine-tune the model performance (Figure 3-4). 
Although a constant setting seemed appropriate for the inflow proportion, the Suwannee 
River outflow proportion varied with processing decade (see model results discussion). 
Flow rates were also varied by vegetation type (shrub, forested, open water, prairie) and a 
manning's coefficient, which affects the flow rate depending on the substrate type (Table 
3-1). Proportional adjustments to the estimated evapotranspiration volumes were also 
included to vary seasonal evapotranspiration rates, if necessary (Figure 3-4, "percent PET 
to use" in the model's user interface). 



03 



217 



&JB g 



CVJCO 



03 
03 








u 

CO 

■a 

c 
JO 

.c 
-*— 

CO 

en 

2 

o 

to 

u. 
<L> 

i 

SO 

c 



o 

B 

c 

I 

■*-» 
CO 

•a 

W 

Q 
O 

I 

O 



•a 

u 

CO 

3 
co 
4> 

c 
o 

N 

o 

CO 

e 

_o 
■+-» 
as 
o 

o 

-J 

■ 

U 

u. 
3 
00 



218 






o 
2 

81 



c 

J5 
a 
-o 
o 

o 

a 

ha 

> 

5 

U 
B 



3 

I 

£ 

o 



c° 



o 

1 

o 

1 



w 

Q 
O 

I 

O 



g 

•a 

<D 
■s. 
3 

<U 
C 

o 

N 

'-~ 

o 

c 
.0 

oi 
o 

o 

vd 

■ 

g 

I 






219 

Table 3-1. Manning's roughness coefficients used in HYDRO-MODEL to adjust surface 
water flow rates over various substrates (adapted from Ward (1996)). 



Substrate or Vegetation Type* 


Manning's Coefficient 


Lawn, Agricultural Field 


0.03 


Bare Ground, Urban Development, Road 


0.025 


Clearcut, Sparse Pine Forest 


0.045 


Upland Forest 


0.08 


Wetland Forest 


0.15 


Wetland Shrub 


0.13 


Wetland Prairie 


0.10 


Submerged Aquatic Vegetation 


0.07 


Open Water, unvegetated 


0.065 


Ditch with grass banks 


0.027 


Impoundment 


0.02 


Canoe Trail, unvegetated peat 


0.07 


Canal 


0.065 


Riverbed, unvegetated 


0.047 


Stream bed, unvegetated 


0.10 



1 Manning's coefficients were assigned to topographic features (e.g., stream, lake) first, 
vegetation type second, where appropriate. 









220 
Model Data Sources 

Origins of HYDRO-MODEL point data sets are discussed in detail in Chapter 2. 
A brief overview of their purposes, sources, and conversion to spatial data sets is 
presented here. 
Precipitation 

Precipitation data were compiled from recording stations installed and 
maintained by NOAA, the Okefenokee National Wildlife Refuge, and supplemental 
recorders installed in this study. Biweekly precipitation totals were calculated and 
missing data were estimated using regression relationships developed as detailed in 
Chapter 2. The biweekly point data were converted to spatial grids using interpolation 
algorithms in ARCINFO. Several methods provided by ARCINFO and ARCGRTD were 
applied to the point data, and a methodology combining the techniques was chosen. Data 
from each biweekly interval were first interpolated with kriging and the circular 
algorithm, selected after viewing interval semivariograms and determining that in 
general, the most realistic surfaces were calculated with this algorithm (Burroughs 1986). 
Where insufficient data densities prohibited using this method, the point data were 
interpolated using the ARCINFO-TINNING command (ESRI 1992) with the quintic 
algorithm, and then gridded to 500x500 m cells. The resultant precipitation data surfaces 
were stored in a directory and accessed individually during model processing. 



221 
Evapotranspiration 

Biweekly evapotranspiration point data were estimated from temperature data 
collected daily at NOAA weather stations, as detailed in Chapter 2. Procedures to 
calculate data surfaces followed those detailed for precipitation data. Evapotranspiration 
estimates were modified with a multiplier to account for differential rates in major 
vegetation types, and a user-defined coefficient was added to the model interface to 
seasonally adjust evapotranspiration rates (Figure 3-4). 
Creek Inflow Volumes 

Surface water flow into the swamp is concentrated along the northwestern 
perimeter (Figure 2-24). Creek inflow volumes at selected locations were estimated as 
detailed in Chapter 2. Volumes representing biweekly flow estimates were converted to 
sheetflow by proportional dispersion across inflow zones delineated using the swamp 
vegetation and topographic maps as guides of zone boundaries (Figure 3-5). Distribution 
of inflows into regions rather than from perimeter points reflected Blood's (1981) 
conclusions that the bifurcation ratio of the northwestern inflowing streams was 
indicative of a low relief, coastal plain where loosely defined stream channels and 
branching are common. This implies that water movement into the swamp can be 
represented as sheetflow rather than as point inflow sources. Therefore, each cell in the 
input zone grid received a proportion of the biweekly, total volume inflowing from 
northwestern perimeter creeks, in Phase I of the model. A user-defined coefficient to 
uniformly modify this proportion was also added to the model interface to facilitate 
adjustment (Figure 3-4). Biweekly inflow data are stored in a data table (IN4193) and 



222 
are applied to the appropriate zone grid as identified in an INFO table item (INFLOW). 
Because groundwater contribution to the total swamp water budget is minimal (Rykiel 
1977), it was not included as a separate model parameter but included in the surface 
inflow volumes. 
River Outflow Volumes 

Volumes of water leaving the refuge via the St. Marys River, Suwannee River, 
Cypress Creek, and Sweetwater Creek were estimated as detailed in Chapter 2. 
Biweekly flows were converted to sheetflow by proportional dispersion across outflow 
zones delineated using the swamp vegetation and topographic maps as guides of zone 
boundaries (Figure 3-6). Each cell in the zone grid received a proportion of the total 
outflowing volume in Phase I of the model. A user-defined coefficient to uniformly 
modify this proportion in the Suwannee River outflow zone was also added to the model 
interface to facilitate adjustment (Figure 3-4). Biweekly outflow data are stored in a data 
table (OUT4193) and are applied to the appropriate zone grid as identified in an INFO 
table item (OUTFLOW). Only the Suwannee River outflow volume was removed in the 
model iterations discussed here. Although the model includes instruction to similarly 
remove outflow from the other exiting flows, the topographic gradient was used to force 
directional flow in these areas. Groundwater outflow was assumed to be a minimal 
component of the total swamp water budget (Rykiel 1977), and therefore it was included 
in the outflow estimate instead of as an independent model parameter. 



223 
Water Depth and Topographic Surfaces 

Direction of water movement through cells in the swamp landscape (outside of 
inflow and outflow zones) is determined by the surface topographic gradient. 
Topography grid development is detailed in Chapter 2. Starting water depth grids 
estimated for the first interval of each decade are retrieved by the model and added to the 
topography surface to create a starting water surface elevation grid (Figure 3-2, Phase I). 
Creation of the water depth grids is detailed in Chapter 2. Movement of water among 
grid cells in the water surface elevation grid is accomplished with a neighborhood query 
and summation (Figures 3-2 and 3-3, Phase 3), and modified with a Manning's 
coefficient (Table 3-1) to allow for differential movement of water across varying 
substrates (Ward 1995). The ending water depth grid is created by subtracting the 
topographic surface elevation grid from the water surface elevation grid, and the 
resultant depth in Phase 3 becomes the starting water depth in Phase 1 of the subsequent 
interval (Figure 3-2). 
Data Surfa ces Used for Model Assessment 

At the end of the completed model run, an assessment of model performance was 
made using water depth estimates extracted from 30 cells corresponding to water level 
recorder locations (Figure 3-7). The subroutine "Check Stations" (Figure 3-2) extracted 
the station name, date, water surface elevation, and water depth for each interval and 
created an ASCII file that could be imported into a spreadsheet program to plot against 
recorder data. The entire water depth, water surface elevation, and water movement 
grids could also be examined using the "Display Results" subroutine (Figure 3-2) to view 



224 



SiMamee 
Creek 



Kingfisher 
Landing 




Cypress 
Creek 















Figure 3-7. Locations of water level recorders used to assess HYDRO-MODEL 
performance. 



225 
spatial relationships among the recorder location data and the surrounding swamp 
landscape. Subsequent adjustments to the model code were based on these visual 
comparisons. 
Model Manipulation and Assessment 

The primary objective of this study was to determine if the sill is affecting the 
swamp hydrologic environment, and if so, to what spatial and temporal extent. 
Manipulations of the model code and swamp topographic surface provided initial 
indications of the sill's impacts. The model was constructed using a topographic surface 
representing the "with-sill" condition during 1980-1993 (Figure 2-20). This surface was 
replaced with a "no-sill" topographic surface (Figure 3-8) and model variables set at 
"pre-sill" levels to estimate the swamp hydrologic environment during 1980-1993 in the 
sill's absence. Similar conditions were set for 1960-1969 and 1970-1979 data and the 
"no-sill" topographic surface to assess possible changes in swamp hydrology that might 
be attributed to the sill. The model was also manipulated with 1941-1949 and 1950-1959 
data. The topographic surface including the sill was used, with model parameters set at 
"with-sill" levels, to approximate conditions that might have existed with the sill in 
place. Model sensitivity to changing water volume was also assessed. In separate model 
runs the total Suwannee River outflow was also retained in the swamp during each 
decade to determine the maximum impoundment levels possible. Additional model 
manipulations included incremental increases and decreases in Suwannee River outflow, 
evapotranspiration volumes, and volumes of creek inflow to assess responses in the 
1980-1993 swamp environment. 



226 













Figure 3-8. Estimated topographic surface representing the pre-sill peat surface 
elevations. Dark areas are low in elevation. 



227 
After each model manipulation the "Check Stations"summary file was created 
and imported into a spreadsheet to graphically compare with other model manipulation 
results. Since the model was constructed using recorder data from 1980-1993, model 
performance was best during that decade. Disagreements between model estimates and 
estimated recorder data for 1941-1979 may be a function of system changes (e.g., 
topography, inflow volumes, vegetation distributions) or missing data extimation 
techniques, and not necessarily indicte a poor model performance. Therefore, model 
manipulations during 1941-1979 were exploratory, while those for 1980-1993 were used 
to identify the sill's influence on the system. Decade and growing/nongrowing season 
hydroperiods were also calculated from "with sill" and "no sill" model results, and 
contingency tables (log-likelihood ratio, G-statistic) (Sokal and Rohlf 1981) were used to 
determine where hydroperiod frequencies differed significantly. Changes in 
relationships between the sill area water depths and those at stations throughout the 
swamp under high, average, and low water level conditions were assessed by comparing 
coefficients of variation and slopes of regression relationships. These assessments 
provided clues to the spatial and temporal extent of the sill's effects, and variability of 
these effects with overall water level conditions. 
Wildfires in the Area Affected bv the Suwannee River Sill 

The primary purposes of the Suwannee River Sill were to facilitate wildfire 
control by creating impounded conditions during periods of drought, and to arrest the 
spread of wildfires across the landscape by prolonging inundation. Refuge records 



228 
contain information primarily on fires that were controlled by fire suppression 
intervention and not on those that were initiated and naturally extinguished before 
detection. Therefore it is not possible to determine if the sill affected total fire 
occurrence. However, it is possible to determine if the sill was elevating water levels 
during seasons of high fire frequency, if fires were arrested in the sill impact area due to 
elevated water levels, and if reported incidences of wildfires decreased following sill 
construction. These questions were addressed by comparing maps of wildfire ignition 
location and burn extent with a delineation of the sill-affected area, and information on 
general hydrologic conditions at the time of the wildfires, summarized from the water 
level recorder database and model output surfaces. Comparisons were made using 
IMAGINE (version 8.2, ERDAS, Inc., Atlanta, GA 30329) summaries and overlays and 
ARCVIEW map inqueries. 
Vegetation in the Area Affected by the Suwannee River Sill 

The Okefenokee Swamp vegetation landscape is dynamic. Fluctuations in 
species compositions and distributions may be the consequences of naturally occurring 
community succession, but may also result from historic logging, wildfire management, 
or manipulations of the landscape hydrology. Comparisons in vegetation distributions 
relative to logging history and wildfire history are detailed in Chapters 4 and 5, 
respectively. Changes that might be attributed to hydrologic modifications of the 
Suwannee River Sill are summarized in this chapter. Areas of vegetation change 
determined in Chapter 4 were compared with ERDAS-IMAGINE summary overlays of 



229 
the estimated sill impact area. Proportions of vegetation types within and outside of the 
affected area were estimated and compared between the areas. 

Results 

Area Affected by the Suwannee River Sill 

Although varying with precipitation and evapotranspiration volume, the northern 
and eastern extent of the Suwannee River sill's effects are roughly delineated by 
Craven's Hammock to Floyd's Prairie to southeastern Chase Prairie to the Pocket (Figure 
3-9). Since sill construction, this region has experienced elevated water levels and/or 
extended hydroperiods that have not occurred elsewhere in the swamp. The sill has also 
affected vegetation composition and distribution in this area, which were previously 
altered by logging and fire suppression. Discussion of changes in the swamp hydrologic 
environment and vegetation distributions indicated by the hydrology model output 
follows. 
Model Accuracy: 1980-1993 

Model performance was assessed at 28 stations distributed throughout the swamp 
(Figure 3-7); model data and trends at 20 of these stations generally followed recorder 
data (Figure 3-10). These 20 stations were used to ascertain effects of additional model 
manipulations. The remaining 8 stations elucidated various problems with the model 
(Figure 3-11). Excessive water depths at perimeter or near-perimeter stations were 



230 



I 

i-fi 



III 

III 



111 




S 
•5 

I 



8-4 



(30 



u 



g 







c 



o p 

+■* T3 

J J 

i- c 

> ' ! 






a 
x 



- — 



1! 

j| 

<~- o 
o •& 

+- 1) 

ll 

° 1 

eo .2 

P • s 

I" 

T3 o 

9 t: 
£ § 

s/3 — 

s ' 

<u 3 

§> s 

tn > 



2.0 i 



£ 1.5- 

5 

a 

| 1.0 

u 

I 0.5 - 



0.0 
2.0 



Billy's Lake 



231 



a 

0) 

a 

u 
(0 



(0 

5 



B 1.5- 



1.0 - 



0.5 - 



0.0 
2.0 



E 1.5 

5 

a 

S i.o 



0.5 - 



0.0 



- Recorder 

With-Sill 




T 



-- No Sill 
— With-Sill 




No Outflow 
With-Sill 




~~ i i i i i i i i i i — i — i — i — i — r~ 
1980 1982 1984 1986 1988 1990 1992 1994 

Year 

Figure 3-10. Estimated recorder data and model output from the "with-sill" and "no-sill' 
simulations for 1980-1993. 





Chase Prairie 232 


1 ^ 






1 .0 






^ 




Recorder 


Su- 




With-Sill 


fi 






a 






0) 






terD 

o 

bi 

i 


/ V\ A 


Kk^tK yL\,^ t\»*fo 


CO 


\/' w 


/ v VW V \7 h X- *8* A-v AVT r 


0.0 - 
1.5 - 






I I I I 










No Sill 


^■^ 






£ - « 




With-Sill 


*•* 1.0 - 






5 






a 






Q> 






2 0.5- 






© 


%t l\ 


/\» /A l« ril U Jt 


-*-> 


Vp/'\ A 


r "\ //••* A n It \ h Jl /I 


CO 

* 0.0 - 
1.5 - 




/ *Hc s\ A i F\. n Jl, i * p\ A ^ (A 

i . ir \ /A A A ' \l t i \ / vj A n \ \m H m//t 
/ •" sv, w VnV -if »\ i ^*\<o\ /l ?l \Ir k ..i 


I I I I 










No Outflow 


<^^» 






E - « 




With-Sill 


*■* 1.0 - 






5 






a 






CD 


. 


> • 


g 0.5- 


Jl 


j ft ' rl 

W Jl JL | t ' 


0) 


^ / V A 


J W / ¥ * II / f ft ' jl 


*-» 






CO 


\ ft 


/ ■ / v jtV • / ■ #i Ii * ' n sL n 


£ « 




" W ^J % / *^J/ V (ill/ \|W 1 fttf 


^ 0.0 - 








I ! I 

1980 1982 


1 1 1 1 1 

1984 1986 1988 1990 1992 1994 




Year 


Fieure3-10~continued. 





Chesser Prairie 



233 



2.0 



I 1.5- 



Q. 

S i.o- 

S 0.5 - 



0.0 
2.0 



£ 1.5- 



a 

S 10- 

L. 

| 0.5 - 



0.0 
2.0 



£ 1.5- 



S 1.0- 
| 0.5 - 



0.0 



Recorder 

With-Sill 




i — i — r 



i — r 



No Sill 
With-Sill 




No Outflow 
With-Sill 




1 1 1 1 1 1 1 \ 1 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued. 



1.5 - 



B 1.0 - 

Q. 
0) 
Q 

fc 0.5 ■ 

S 

0.0 - 



1.5 - 



S= 1.0 - 

CL 

o 

fc 0.5 - 

(0 

S o.o 



1.5 



£ 1.0 

Q. 
0) 

O 



2 

i 



0.5 - 



0.0 - 



Coffee Bay 



234 



Recorder 
With-Sill 




i — i — r 



i r 



i — i — i — r 



No Sill 
With-Sill 




i — r 



t — r 













No Outflow 




m 


it 


■ 

4 




With-Sill 

• 




t fk. } 


JL 




• 


H 


K A 2 J 








•\ 


* * Tv 




V \7 \ 


f Vi 




IV. i 


i \ft jfV I V 


i i i i 


I 1 


i 


1 


1 1 1 


1 1 1 1 



1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Fi gure 3-10-continued. 



2.0 
1 1.5 

JZ 

*-» 
a 

2 

| 0.5 - 



Craven's Hammock 



235 



0.0 
2.0 



E 1.5 



a 

& 10 H 

| 0.5 - 



0.0 
2.0 



I 1.5- 



a 



| 0.5 - 



0.0 



• - - Recorder 
With-Sill 




I • » * >, 

• • 'l )••'■■ 



T 



No Sill 
With-Sill 




No Outflow 
With-Sill 




AA ;\ .'• 
j t * ■ i 



I I I I I I I I I I I I I — I — 
1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3- 10~CQPtiiw?(l, 



9 n 




Cypress Creek 


23 


z.u 






Recorder 


^s 








E 1.5- 






With-Sill 


a 








Q. 




•:• • p Y a 


.i * l\ 


& 10- 




."»• ■]» ' 7 "YAP ■• V- 1 ' U J 




4S 0.5 - 






\ I ■ * 1 * l .a 

\ r *• -V 

A.I • • 


^ 




J\ * ! • *"v 


TV 






i i'V 


«\ ,x 


n n 








U.U *"| 


1 1 


i i i i i i i i i 


i i i i 



2.0 



£, 1.5 - 

5 

a 

& 10 

| 0.5 



0.0 
2.0 



E 1.5- 

a 

a 

| 0.5 



0.0 



No Sill 
With-Sill 




A. A *v '. 



i — T~^ i — I — I — i — i — i — 1 — f — i — i i 



No Outflow 
With-Sill 




"~ i i i r i i r 
1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3- 10~contipu(?(3, 



1.5 



5 

a 



2 0.5 

3 

(0 

* 0.0 - 



1.5 



a 

? 0.5 
5 

(0 

0.0 - 



1.5 



1 1.0 

*-» 
a 
s 
Q 



& 



0.5 



0.0 - 



Double Lakes 



237 



i — i — r 



i — r 



i r 



Recorder 

With-Sill 




T 



No Sill 
With-Sill 




i i i i r 



No Outflow 
With-Sill 




I I I I I I I I I I I 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3- IP-continued 



2.0 



E 


1.5 


■*-> 
a 

0) 
Q 


1.0 


5 

CO 


0.5 




0.0 




2.0 


? 


1.5 


a 

0) 

a 


1.0 


(0 


0.5 




0.0 




2.0 



E 1.5- 



® 1.0 

L. 

| 0.5 - 



0.0 



Durdin Prairie 



238 



Recorder 

With-Sill 




T 



No Sill 
With-Sill 




i — i — r 



No Outflow 
With-Sill 




I I I I I I 1 I I I I I I I 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued. 

















Floyd's Prairie 


239 


1.5 - 


















Recorder 




^"*. 








£ - A 




With-SHI 




^ 1.0 - 








£ 








a 








V 




* 1 




2 0.5- 
& 


y* \ A / l r i i ' /A J i A 

j * * \ / •< / v.w" vi ri •••li I > 1 1 ii\ i A A 


U A 1 /J 
*J * *\ b \ imi " i 

f ■■' \rv \ i 




CO 


•. . sB v v\/ n/ "V ••^vin 


1 f v*/ 




5 0.0 - 
1.5 - 




w 




1 1 1 1 1 1 1 1 1 1 1 1 










No Sill 




<■» * 








E 




With-Sill 




« 1.0 - 








5 








a 


1 «V /\ A A R 






0) 


/ i w^ '\ 11 1 l.i H 






2 0.5- 


/ ■ 'A t Nfiii « Lit » 

• \ /.^S i '. V' M A rl t v\ f\i A 1 


/^'A / "1/11/ "'l 




(0 

* 0.0 - 
1 => 


I ■ * » ». i J i LI .11 >*r ,. 1 *\ f * 






i i i i i i i i i i i 


1 .O 


m 


- - - No Outflow 




^1 1^ 








E 


t 


With-Sill 




* 1.0 - 


« •• • > ■ 1 1 ^ i 


1 




r Depth 

o 

i 


. V I < ■■ ■ 1 • 

N\ .'A'. ft -'V » * ,• .' \ 

\v/\ • •/ 1/ \ a " 71 '\ .7 V Ml Vm ' ••' 


• 

i /• 

•'. * ' 

flu V-7\ a, 1 li 




q 


\7 S w V i/\'.'l-7 M\i A' M- 


i Vi W ^ i * 




43 

a 


U \/ \ 7 V7 W VlAVl* 
* \ / vV I ii 1/ H * ■ 1 ■ 


U 1 1 •* 




$ 


» ^w \[ " lu \ »V 






0.0 - 








1 1 1 1 1 i 1 1 1 1 1 
1980 1982 1984 1986 1988 1990 


1 ! 1 

1992 1994 




Year 




Figure 3-10-c 


ontinued. 









Gannett Lake 


24 


9 n 








C..KJ 














- — Recorder 


^m^. 








£, 1.5 - 






With-Sill 


£ 








a 








£ 1.0- 








u 








V 


iA \ 1 \ 


Mm / '• v / \ *'l il In [1 > A 


AX' A * « 


| 0.5 - 


V \ * \ / \ 


/ S\pv\/ V'7 '\/ A Ma A A 


P V \ •! 




^ r 


■ » * 1*1 '1 


kTV *^ 


n n 








u.u 


i i i 


1 1 1 1 1 1 1 


1 1 1 1 



2.0 



I 1.5- 

5 

a 

g 1.0 

L. 
J 

| 0.5 



0.0 
2.0 



I 1.5- 

a 

| 1.0- 

| 0.5 - 



0.0 



1 — i — r 



i — i — i — r 



No Sill 
With-Sill 




No Outflow 
With-Sill 




\ I I I I I I I I I I 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued 



1.5 



*1jOH 

a 

O 

o 



5 

(0 



0.5 



0.0 - 



1.5 



£ 

a 

Z 0.5- 

o.o -I 

1.5 



I 1.0- 



a 
a 
a 

| 

(0 



0.5 - 



0.0 - 



i r 



Moonshine Ridge 



241 



i — r 



Recorder 

With-Sill 




No Sill 
With-Sill 




1 r 



i i r 



No Outflow 
With-Sill 




I 1 I I I I I I I 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued 



1.5 



Slo 



Sapling Prairie 



242 



a 
o 

2 

CO 



0.5 - 

0.0 
1.5 



£1.0 

f 

0) 

I 0.5 

3 

(0 

* 0.0 



1.5 



£1.0 

*■* 
a 

? 0.5 
I 

* 0.0 



Recorder 

With-Sill 




i i i i r 



No Sill 
With-Sill 




No Outflow 
With-Sill 




I 1 I I 1 I I I I I I I 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-l0"CQntinued, 



2.0 



E 1.5- 



| 1.0 



3 0.5 

CO 

0.0 



2.0 



E 1.5 - 



§■ 1-0 



2 0.5 

(0 

0.0 



2.0 



E 1.5- 



a 


1.0 


Q 




l. 




.2 


0.5 


CO 




£ 






0.0 



Sapp Prairie 



243 



Recorder 

With-Sill 




No Sill 
With-Sill 




i — i — r 



No Outflow 
With-Sill 




1 I I I I I I I I I I I 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued. 










SCFSP 244 


2.0 - 


Recorder 




? 

£ 1.5 - 
a 

a 

Q 1.0 - 

| 0.5- 
n n 


With-Sill 




u.u 

2.0 - 

? 

r 1.5 - 

a 

° 1.0 - 

.8 

| 0.5- 

nn-i 


i i i i i i i i i i i i i i i 




No Sill 


» With-Sill 

'J' V Ar V'' V W''Vi' ' \/aV \ /A K V i A * 

V V/ V V W \j V vv v ^y v ■ » y 




u.u 

2.0 - 

? 

T 1.5 - 

a 
o 

Q 1.0 - 

1 

| 0.5 - 


1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 


No Outflow 

.\ ;. t With-Sill 

• « .• * ■• 

7\ . / * i« i V ' V . /• .* /> '. /' 

\ , v V v y V V V V /• ' V ■ 

i ■ ^ \l V * • » i | V • 


n n 






uu i i i i i i i i i i i i i i i i 
1980 1982 1984 1986 1988 1990 1992 1994 




Year 


Fieure 3-10-continued. 









1.5 



*10H 

5 
a 


? 0.5- 



3 

(0 



0.0 



1.5 



f 1.0H 

a 
a 

? 0.5- 

> 0.0 - 



1.5 



§1.0- 



a 

CD 

a 
I 



0.5 - 



0.0 - 



i r 



SCRA 



245 



i — i — i — i — r 



i r 



i r 



Recorder 

With-Sili 




i — r 



No Sill 
With-Sill 




No Outflow 
With-Sill 




I I I I I I I I 1 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued. 



Sill Area (South Gate) 



246 




| 2.0 H 

f 1.5- 

5 

Q 

i 10 

*-• 

(0 

£ 0.5 - 



0.0 



- No Outflow 
— With-Sill 



;• a •■■ a a »•■ . 

K a ^ v v v v v\ a 




I I I I I I I I I I I — I — I — I — I — 
1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-10-continued 









Suwannee River 




24 


G. 












Z.D 























Recorder 


f 2.0- 










- With-Sill 














e 












U 1-5 - 






■ 






0) 






•i 


i 




Q 


■ r 






l", '* 


■ 


fe 1.0 - 


Xr 




y\ a < A 71 » 


1 I 

l A A 




1 

(0 




\ / N 




\ 1 i \ 


A *>n 


£ 0.5 - 


i 


* 1 .• •. 




TfW, 








V 1 


\ t • «. v 


'«« 


£ 


nn - 












u.u 


1 1 


1 


1 ! 1 1 1 1 1 1 


1 1 


1 1 



2.5 



2.0 - 



1 1.5 - 
a 

Q 

.- 1.0 - 

i 

£ 0.5 - 



0.0 
2.5 



|2.(H 

f. 1.5 

0) 
Q 

|1-0H 

(0 

£ 0.5 - 



0.0 



- No Sill 
- With-Sill 




v u v v Sf \- v * vj vy v •/ 



T 



- No Outflow 

- With-Sill 



f\!"J\ 




i i i i i i i i — i — i — i — i — i — i — i — 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3- 1 0-continued. 



1.5 



* MM 

3 

a 
o 

Q 



1 

CO 

5 



0.5 - 

0.0 - 
1.5 



&1JH 

5 

a 

0) 

2 0.5- 

(C 

* 0.0 - 



1.5 



tl.0H 



a 

Q 
(0 

5 



0.5 - 



0.0 - 



Sweetwater Creek 



248 



Recorder 
With-Sill 




i — r 



No Sill 
With-Sill 




i — r 



No Outflow 
With-Sill 




I I I 1 I I 1 1 1 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 



Year 












Figure 3- KbfiflBtinuaL 



1.5 



I 1.0- 



a 
© 
a 

3 

CO 



0.5 - 

0.0 - 
1.5 



f 1.0H 

53 

a 
3 

2 0.5 

.8 

(0 

* 0.0 - 



1.5 



*10 



a 

0) 

a 

(0 



0.5 - 



0.0 - 



Territory Prairie 



249 



Recorder 
With-Sill 




i r 



i r 



T 



No Sill 
With-Sill 




i — i — i — r 



i r 



No Outflow 
With-Sill 




I I I I I I I I I I I 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figyn? 3-10~cQirtipwfl, 



Transect 60 



250 



2.5 



? 


2.0 


^— " 






1.5 


a 




a> 




O 


1.0 


L. 









4-* 

(0 


0.5 


^ 


0.0 



2.5 



? 


2.0 


■«-^ 




*-» 


1.5 


a 




a> 




Q 


1.0 


V- 




0) 




(0 


0.5 


£ 






0.0 



2.5 



? 


2.0 


^"^ 




x: 


1.5 


a 




0) 




a 


1.0 


b. 




a) 




to 


0.5 


£ 






0.0 



Recorder 

With-Sill 




T 



No Sill 
With-Sill 




/\ a ••. A ;. \ . 

v \. v V\ A 



No Outflow 
With-Sill 



V 



is' V 
V 



4 
/• 
■ i 




I I I I I I I I I I 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3- 1 0-continued 



2.0 

? 
~ 1.5 

9 

a 
S 10 

* 0.5 

0.0 - 



2.0 - 
? 

r 1.5 



S i.o 

(1) 

15 0.5 



0.0 - 



2.0 - 



E 




.c 


1.5 


■*-> 




a 




0) 

a 


1.0 


i_ 




0) 




*-> 
(0 


0.5 



0.0 - 



Brown Trail (NE of Sill) 



251 



Recorder 
With-Sill 




"i — r 



No Sill 
With-Sill 







No Outflow 
With-Sill 






i\ 



i \ * 
. '.i 



f. 

t l 






V \ 



A ' vv 






V 




I I I I I I I — I — I — I — I — I — I — I — I — 
1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-11. Estimated recorder data and model output from stations with poor model 
performance in "with-sill" and "no-sill" simulations for 1980-1993. 



2.0 



Honey Prairie 



252 



E 


1.5 


JZ 




■*-> 




a 




Q 


1.0 


i- 




i 




(0 


0.5 




0.0 




2.0 


<< ^ > 




E, 


1.5 


x: 




-•-j 




a 




0) 

a 


1.0 


i- 




0) 




*-» 




(0 


0.5 




0.0 




2.0 




I 1.5- 

f 

g 10 

| 0.5 



0.0 



No Outflow 

With-Sill 




I I I I I I I 1 1 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-11 -continued. 



1.5 



Ii.o 



a 
o 

1 



0.5 - 

0.0 
1.5 



SlJH 

t 

Q 

Q 



5 
(0 



0.5 - 



0.0 - 



1.5 



£1.0- 



a 
O 

1 

CO 



0.5 



0.0 



Kingfisher Landing 



253 



- Recorder 
— With-Sill 




i — r 



l r 



No Sill 
With-Sill 




■S3 V 



i r 



i — i — r 



i — r 



- No Outflow 
- With-Sill 




\J V 



I 1 I I I I 1 I I I I I 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-11 -continued. 



a 

Q 

J 

CO 



a 
o> 
a 

s. 

a) 

CO 



a 
o 

CD 

■*-> 

(0 

5 



2.5 
2.0 - 
1.5 - 
1.0 - 
0.5 
0.0 H 
2.5 
2.0 - 
1.5 - 
1.0 - 
0.5 
0.0 
2.5 
2.0 - 
1.5 - 
1.0 
0.5 - 
0.0 - 



Seagrove Lake 



254 



Recorder 

With-Sill 










i — r 



An/* 



1 — I — i — i — r 



No Sill 
With-Sill 




i i — i — r 



No Outflow 
With-Sill 




I 1 I I I I I I 1 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-11 -continued 



2.0 



Soldier's Camp 



255 



£ 1.5 - 



5 

a 

a 



1.0 



5 

I 0.5 



0.0 
2.0 



E 1.5 

£ 

a 

| 1.0 

3 

| 0.5 



0.0 
2.0 



§ 1.5 



S i.o- 

u 
.2 
| 0.5 - 



0.0 



Recorder 

With-Sill 




1 — r 



No Sill 
With-Sill 




i — i — r 



i — r 



No Outflow 
With-Sill 




1 1 1 1 1 1 I 1 1 1 1 I I I I 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-1 l-contipued, 



1.5 



f 



e 0.5 



Suwannee Creek 



256 



a 

CO 



0.0 - 
1.5 



4l0H 

2 0.5- 

to 

^ 0.0 - 



1.5 



S1.0 

I 

a 
2 0.5 

i 

0.0 



- Recorder 
— With-Sill 




i — r 




No Sill 
With-Sill 



TLA-^JlA j . L 



.k • «& 



i — r 



No Outflow 

With-Sill 




i 



n i i i i i i i i i — i — i — i — i — i — 
1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-1 l-continued 



Transect 52 



257 



E 




w * 


1.5 


c 




*■> 




a 




Q) 
Q 


1.0 


i- 




(1) 




(0 


0.5 



2.0 - 
? 

4-) 

a 
g 1-0 

%m 

1 0.5 

0.0 



2.0 - 

? 

IT 15 

a 

S i.o 

Li 

I 0.5 



0.0 - 



2.0 



0.0 - 



Recorder 
- With-Sill 








No Sill 






With-Sill 


U\ A A A * 

I I I I I I 


A A A 

i i i i 


. IS a A 

I I I I I 



No Outflow 
With-Sill 



' J '• . \ \ ". • \ ' » ' ' ■ 
A, V * * v V V v,A 



A 




V 



. v '. 



JVa_A_ 



i i i i i i — i — i — i — i — i — i — i — i — i — 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-11 -continued. 



Transect 55 



258 



2.0 - 



E 




]T 


1.5 


^-* 




a 




0) 


1.0 


u 




0) 




(0 


0.5 



0.0 - 



2.0 
? 

a 

S i.o 

u 

a) 

« 0.5 



0.0 - 



2.0 - 



E 




^** 


1.5 


JL 




-m 




a 




0) 

Q 


1.0 


i_ 




0) 




—> 
CO 


0.5 



0.0 - 



Recorder 

With-Sill 




!•• » k *i , 

1 ' r .'* : 



i — r 



i — r 



i — r 



i — r 



No Sill 
With-Sill 




- - No Outflow 

— With-Sill 




I I 1 1 1 1 1 1 1 1 1 1 1 1 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-11 -continued. 



259 
that reached the perimeter was forced back into the swamp unless in the outflow zone or 
in areas with topographic data beyond the refuge perimeter (South and Southwest). This 
probably accelerated accumulation at these sites (Seagrove Lake, Soldier's Camp). 
Additionally the paucity of topographic survey points along the eastern edge leading up 
to Trail Ridge may have contributed error to the topographic surface and subsequently to 
the water depth calculations (Figure 2-18). Poor agreement along the east-central 
perimeter may also reflect the lack of information about the hydrology of the seepage 
flows entering the swamp in this region. The model may have underestimated water 
depths at Kingfisher Landing, Chase Prairie, Sill, Honey Prairie, and Transect 52 and 
Transect 55 because the topographic environments at these recorder locations were 
atypical of the area and these features were not preserved in the water depth 
interpolations. This is most likely an artifact of model data scale, which was limited by 
computer resources. Model output was more often in agreement with water level 
recorder data that had been interpolated among stations using kriging and a 50m cell 
size, than data extracted directly from the recorders. Non-kriged data were in better 
agreement with model output than interpolated data where recorder locations were more 
representative of the area's general topography, and stations were not located in trails, 
canals, ditches, or holes in the peat (Table 3-2). The model appeared to suitably 
represent outflowing creeks and rivers using the Suwannee River outflow zone and 
topographic gradients, but water depth estimates at the Suwannee Creek recorder site 
were low. Adjustments to the inflow and outflow proportion coefficients did not 
improve this performance. 



260 



Table 3-2. Best HYDRO-MODEL settings and check data format for stations in Okefenokee 
Swamp during 1941-1993 model simulations. 



Station and Interval 


Evapotranspiration 
Coefficients" 


Suwannee 

River 

Outflow 

Coefficient 


Check Data 
Type 


Model 
Performance 


1941-1949 










Billy's Lake 


1.0,1.0,1.0 


0.20 


kriged 


fair 


Chase Prairie 


1.0, 1.0, 1.0 


0.20 


non-kriged 


fair 


Chesser Prairie 


1.2,1.2,1.2 


0.30 


non-kriged 


high 


Coffee Bay 


1.2, 1.2, 1.2 


0.30 


kriged 


fair 


Craven's Hammock 


1.2,1.2,1.2 


0.30 


non-kriged 


fair 


Cypress Creek 


1.15,1.15,1.15 


0.20 


non-kriged 


fair 


Double Lakes 


1.15,1.15,1.15 


0.20 


kriged 


fan- 


Durdin Prairie 


1.2,1.2,1.2 


0.30 


kriged 


fair 


Floyd's Prairie 


1.15,1.15,1.15 


0.20 


non-kriged 


fan- 


Gannett Lake 


1.2,1.2,1.2 


0.30 


kriged 


high 


Honey Prairie 


1.0,1.0,1.0 


0.20 


kriged 


fair 


Kingfisher Landing 


1.0,1.0,1.0 


0.20 


kriged 


fair 


Moonshine Ridge 


1.2,1.2,1.2 


0.30 


kriged 


fair 


Suwannee River 


1.0, 1.0, 1.0 


0.20 


kriged 


fair 


Sapling Prairie 


1.2, 1.2, 1.2 


0.30 


non-kriged 


fair 


Sapp Prairie 


1.0,1.0,1.0 


0.20 


kriged 


low 


SCFSP 


1.0,1.0,1.0 


0.20 


kriged 


fair 


SCRA 


1.2,1.2,1.2 


0.30 


kriged 


fair 


Seagrove Lake 


1.2,1.2,1.2 


0.30 


kriged 


high 


Brown Trail (Sill) 


1.0,1.0,1.0 


0.20 


kriged 


low 


South Sill Gate 


1.0,1.0,1.0 


0.20 


kriged 


fan- 


Soldier's Camp 


1.2, 1.2, 1.2 


0.30 


kriged 


high 



Table 3-2 -continued. 



261 



Station and Interval 


Evapotranspiration 
Coefficients* 


Suwannee 

River 

Outflow 

Coefficient 


Check Data 
Type 


Model 
Performance 


Suwannee Creek 


1.0,1.0,1.0 


0.20 


non-kriged 


low 


Sweetwater Creek 


1.0, 1.0, 1., 


0.20 


kriged 


fair 


Territory Prairie 


1.15, 1.15, 1.15 


0.20 


kriged 


fair 


Transect 60 


1.2,1.2,1.2 


0.30 


kriged 


fair 


1950-1959 










Billy's Lake 


1.25, 1.25, 1.25 


0.20 


kriged 


good 


Chase Prairie 


1.25, 1.25, 1.25 


0.15 


non-kriged 


good 


Chesser Prairie 


1.25,1.25,1.25 


0.20 


non-kriged 


high 


Coffee Bay 


1.25, 1.25, 1.25 


0.20 


kriged 


high 


Craven's Hammock 


1.25, 1.25, 1.25 


0.15 


kriged 


fair 


Cypress Creek 


1.25,1.25,1.25 


0.20 


non-kriged 


high 


Double Lakes 


1.25,1.25,1.25 


0.15 


kriged 


fair 


Durdin Prairie 


1.25, 1.25, 1.25 


0.20 


kriged 


fair 


Floyd's Prairie 


1.25, 1.25, 1.25 


0.20 


non-kriged 


good 


Gannett Lake 


1.25, 1.25, 1.25 


0.20 


kriged 


high 


Honey Prairie 


1.25, 1.25, 1.25 


0.15 


kriged 


good 


Kingfisher Landing 


1.25, 1.25, 1.25 


0.15 


kriged 


low 


Moonshine Ridge 


1.25, 1.25, 1.25 


0.20 


kriged 


fair 


Suwannee River 


1.25, 1.25, 1.25 


0.20 


kriged 


fair 


Sapling Prairie 


1.25,1.25,1.25 


0.15 


non-kriged 


fair 


Sapp Prairie 


1.25, 1.25, 1.25 


0.15 


kriged 


low 


SCFSP 


1.25, 1.25, 1.25 


0.20 


kriged 


fair 


SCRA 


1.25, 1.25, 1.25 


0.20 


kriged 


high 


Seagrove Lake 


1.25,1.25,1.25 


0.20 


kriged 


high 


Brown Trail (Sill) 


1.25,1.25,1.25 


0.15 


kriged 


low 



Table 3-2 --continued. 



262 



Station and Interval 


Evapotranspiration 
Coefficients* 


Suwannee 

River 

Outflow 

Coefficient 


Check Data 
Type 


Model 
Performance 


South Sill Gate 


1.25, 1.25, 1.25 


0.20 


kriged 


fair 


Soldier's Camp 


1.25,1.25,1.25 


0.20 


kriged 


high 


Suwannee Creek 


1.25,1.25,1.25 


0.15 


kriged 


low 


Sweetwater Creek 


1.25, 1.25, 1.25 


0.15 


kriged 


fair 


Territory Prairie 


1.25, 1.25, 1.25 


0.15 


kriged 


fair 


Transect 60 


1.25, 1.25, 1.25 


0.20 


kriged 


high 


1960-1969 










Billy's Lake 


1.25,1.25,1.25 


0.15 


kriged 


fan- 


Chase Prairie 


1.25, 1.25, 1.25 


0.15 


non-kriged 


fair 


Chesser Prairie 


1.25,1.25,1.25 


0.20 


non-kriged 


high 


Coffee Bay 


1.25,1.25,1.25 


0.20 


kriged 


high 


Craven's Hammock 


1.25,1.25,1.25 


0.20 


non-kriged 


good 


Cypress Creek 


1.25,1.25,1.25 


0.15 


kriged 


low 


Double Lakes 


1.25,1.25,1.25 


0.20 


non-kriged 


fair 


Durdin Prairie 


1.25,1.25,1.25 


0.20 


kriged 


good 


Floyd's Prairie 


1.25, 1.25, 1.25 


0.15 


non-kriged 


good 


Gannett Lake 


1.25,1.25,1.25 


0.20 


kriged 


high 


Honey Prairie 


1.25, 1.25, 1.25 


0.15 


kriged 


low 


Kingfisher Landing 


1.25, 1.25, 1.25 


0.15 


kriged 


low 


Moonshine Ridge 


1.25, 1.25, 1.25 


0.20 


kriged 


fair 


Suwannee River 


1.25, 1.25, 1.25 


0.15 


kriged 


good 


Sapling Prairie 


1.25, 1.25, 1.25 


0.15 


non-kriged 


fan- 


Sapp Prairie 


1.25, 1.25, 1.25 


0.15 


kriged 


low 


SCFSP 


1.25, 1.25, 1.25 


0.15 


kriged 


good 


SCRA 






kriged 


good 





Table 3-2~continued. 










263 




Station and Interval 


Evapotranspiration 
Coefficients" 


Suwannee 

River 

Outflow 

Coefficient 


Check Data 
Type 


Model 
Performance 




Seagrove Lake 


1.25, 1.25, 1.25 


0.20 


kriged 


high 




Brown Trail (Sill) 


1.25,1.25,1.25 


0.15 


kriged 


low 






South Sill Gate 


1.25,1.25,1.25 


0.15 


kriged 


fair 






Soldier's Camp 


1.25, 1.25, 1.25 


0.20 


kriged 


high 






Suwannee Creek 


1.25, 1.25, 1.25 


0.15 


non-kriged 


low 






Sweetwater Creek 


1.25, 1.25, 1.25 


0.15 


kriged 


high 






Territory Prairie 


1.25, 1.25, 1.25 


0.20 


non-kriged 


fan- 






Transect 60 


1.25, 1.25, 1.25 


0.15 


kriged 


good 






1970-1979 














Billy's Lake 


1.20,1.20,1.20 


0.15 


kriged 


fan- 






Chase Prairie 


1.20,1.20,1.20 


0.15 


non-kriged 


good 






Chesser Prairie 


1.30, 1.30, 1.30 


0.15 


non-kriged 


high 






Coffee Bay 


1.30, 1.30, 1.30 


0.15 


kriged 


high 






Craven's Hammock 


1.25,1.25,1.25 


0.20 


non-kriged 


good 






Cypress Creek 


1.30, 1.30, 1.30 


0.15 


kriged 


good 






Double Lakes 


1.25, 1.25, 1.25 


0.20 


non-kriged 


fan- 






Durdin Prairie 


1.20, 1.20, 1.20 


0.15 


kriged 


good 






Floyd's Prairie 


1.30,1.30,1.30 


0.15 


non-kriged 


good 






Gannett Lake 


1.30,1.30,1.30 


0.15 


kriged 


high 






Honey Prairie 


1.20, 1.20, 1.20 


0.15 


kriged 


good 






Kingfisher Landing 


1.20, 1.20, 1.20 


0.15 


kriged 


low 






Moonshine Ridge 


1.30, 1.30, 1.30 


0.15 


kriged 


fan- 






Suwannee River 


1.20,1.20,1.20 


0.15 


kriged 


good 






Sapling Prairie 


1.20,1.20,1.20 


0.15 


non-kriged 


good 






Sapp Prairie 


1.20, 1.20, 1.20 


0.15 


kriged 


high 







Table 3-2-continued. 










264 




Station and Interval 


Evapotranspiration 
Coefficients* 


Suwannee 

River 

Outflow 

Coefficient 


Check Data 
Type 


Model 
Performance 




SCFSP 


1.20, 1.20, 1.20 


0.15 


kriged 


fair 




SCRA 


1.30,1.30,1.30 


0.15 


kriged 


fair 






Seagrove Lake 


1.30, 1.30, 1.30 


0.15 


kriged 


high 






Brown Trail (Sill) 


1.20, 1.20, 1.20 


0.15 


kriged 


low 






South Sill Gate 


1.20, 1.20, 1.20 


0.15 


kriged 


fair 






Soldier's Camp 


1.30, 1.30, 1.30 


0.15 


kriged 


high 






Suwannee Creek 


1.20, 1.20, 1.20 


0.15 


non- 
mkriged 


low 






Sweetwater Creek 


1.25, 1.25, 1.25 


0.20 


kriged 


good 






Territory Prairie 


1.30, 1.30, 1.30 


0.15 


non-kriged 


fan- 






Transect 60 


1.25, 1.25, 1.25 


0.20 


kriged 


good 






1980-1993 














Billy's Lake 


1.25,1.25,1.25 


0.10 


kriged 


good 






Chase Prairie 


1.25, 1.00, 1.00 


0.15 


non-kriged 


fair 






Chesser Prairie 


1.25, 1.25, 1.25 


0.10 


non-kriged 


good 






Coffee Bay 


1.25, 1.25, 1.25 


0.10 


kriged 


good 






Craven's Hammock 


1.25, 1.25, 1.25 


0.10 


kriged 


good 






Cypress Creek 


1.25, 1.00, 1.00 


0.15 


kriged 


good 






Double Lakes 


1.25, 1.25, 1.25 


0.10 


non-kriged 


good 






Durdin Prairie 


1.25, 1.25, 1.25 


0.10 


kriged 


good 






Floyd's Prairie 


1.25,1.00,1.00 


0.15 


non-kriged 


good 






Gannett Lake 


1.25, 1.25, 1.25 


0.10 


kriged 


good 






Honey Prairie 


1.25,1.00,1.00 


0.15 


kriged 


low 






Kingfisher Landing 


1.25, 1.00, 1.00 


0.15 


kriged 


low 






Moonshine Ridge 


1.25, 1.25, 1.25 


0.10 


kriged 


good 





Table 3-2 --continued. 



265 



Station and Interval 


Evapotranspiration 
Coefficients* 


Suwannee 

River 

Outflow 

Coefficient 


Check Data 
Type 


Model 
Performance 


Suwannee River 


1.25,1.00,1.00 


0.15 


kriged 


good 


Sapling Prairie 


1.25, 1.00, 1.00 


0.15 


non-kriged 


fair 


Sapp Prairie 


1.25,1.00,1.00 


0.15 


kriged 


good 


SCFSP 


1.25, 1.25, 1.25 


0.10 


kriged 


good 


SCRA 


1.25,1.25,1.25 


0.10 


kriged 


good 


Seagrove Lake 


1.25, 1.25, 1.25 


0.10 


kriged 


high 


Brown Trail (Sill) 


1.25, 1.25, 1.25 


0.10 


kriged 


low 


South Sill Gate 


1.25,1.25,1.25 


0.10 


kriged 


good 


Soldier's Camp 


1.25,1.25,1.25 


0.10 


kriged 


high 


Suwannee Creek 


1.25,1.00,1.00 


0.15 


non-kriged 


low 


Sweetwater Creek 


1.25,1.25,1.25 


0.10 


non-kriged 


good 


Territory Prairie 


1.25, 1.00, 1.00 


0.15 


kriged 


good 


Transect 60 


1.25,1.00,1.00 


0.15 


kriged 


good 



a Evapotranspiration coefficients for 4 seasons: April-May and October-November, June- 
September, December-March. 

b Model performance was assessed by visual inspection of agreement between 
hydrographs of model output and check station recorder data. See Figures 3-9 and 3-10 
for model and recorder data plots. 



266 
A single model did not adequately represent the hydrologic environment of the 
entire swamp during 1980-1993. Various settings of the model parameters were tried, 
and the best agreement with recorder data was achieved with 2 versions of the model. 
The swamp basins discussed in Chapter 2 correspond to the areas affected by the 
different models. The variability in model performance reflects the spatial variability of 
the swamp hydrologic environment; spatial overlap in the model reponses is also 
attributed to model processing scale. 

Five stations in the swamp's western basin (Floyd's Prairie, Suwannee River, 
Sapling Prairie, Transect 60, Cypress Creek) demonstrated better agreement with model 
settings providing more surface outflow in the Suwannee River outflow zone 
(setting=0.15) and less ET (settings=1.25, 1.0, 1.0); agreement with recorder data at 3 
stations in the central (Chase Prairie, Territory Prairie) and southwest (Sapp Prairie) 
basins was also best with these settings. An alternative model (ET=1.25, 1.25, 1.25; 
outflow zone=0. 10) with lower outflow volumes and higher evapotranspiration volumes 
agreed with model data at 4 stations in the swamp western basin (Billys Lake, SCFSP, 
Sill Gate, Cravens Hammock, Sweetwater Creek), 2 stations in the northeastern basin 
(Double Lakes, Durdin Prairie), 1 station in the southeastern basin (Moonshine Ridge), 
and 4 stations in the central basin (Chesser Prairie, Coffee Bay, Gannett Lake, SCRA). 

The model reflects the periodicity of water level fluctuations at all of the recorder 
check stations, with cycles of high and low water depths mirroring seasonal fluctuations 
in evapotranspiration. Amplitudes of these fluctuations are less accurate, with model 
output in western and southwestern areas less variable than recorder data, and model 



267 
output in eastern, central, and southeastern areas slightly more variable than recorder 
data. Amplitudes were most accurate at Floyd's Prairie, Sill Gate, Transect 60, Coffee 
Bay, SCRA, and Territory Prairie. Greatest water level fluctuations during 1980-1993 
were recorded in the Suwannee River floodplain (Billy's Lake, SCFSP, Suwannee River, 
Craven's Hammock, Sill Gate, Transect 60), where model error ranged 1-11% of total 
station variability. Water level fluctuations were least in prairie, lake, and canal areas 
(Chase Prairie, Double Lakes, Durdin Prairie, Gannett Lake, Moonshine Ridge, SCRA), 
with model error ranging 5-16% of total station variability. Model error was proportional 
to a site's overall data water depth range; model error was <15% of the recorded range in 
water depth at 17 of the 21 check stations, and <10% at 13 of the 21 check stations 
(Table 3-3). Therefore, model performance was generally sufficient to indicate affects of 
the sill. 
Model Responses to Sill Manipulations 

To approximate the hydrologic environment that might have occurred had the sill 
been absent during 1980-1993, the model was modified to use the pre-sill topographic 
surface. The flow rate in the Suwannee River was set to 0.20, similar to that used in the 
"no-sill" model runs of 1941-1959, and evapotranspiration rates (1.25, 1.25, 1.25) were 
similar to those in the 1980-1993 "with-sill" model runs. Biweekly changes in water 
depths at creek stations are illustrated in Figure 3-10, and differences from "with-sill" 
averages are listed in Table 3-4. Greatest changes in water depths were measured at 
Cypress Creek, SCFSP, Billy's Lake, Sapp Prairie, Suwannee River, and the Sill Gate 






268 



m 

ON 

as 
© 

00 
On 



3 

a 

■*-» 

3 
O 

"5 

O 

E 

+■• 
vi 
1> 
JD 

"O 

c 

I 

c 
o 

i 

t» 

_*: 
o 

<D 

JS 
O 
— 
O 

B 
O 

VI 

"C 

CO 

a, 

E 
o 
U 

■ 

I 





m 


F 






































■ 






































s 

E 


1 


2 




o 


o 




o 




o 


© 




o 




o 




o 




o 




H 


A 


VI 




































fa 


■e 


E 




































£ 


1 


u 

o 




o 


o 




o 




o 


o 




OO 




o 




o 




o 




H 


X 


'v, 




































If 

u A 


1 
a 


1 

VI 


E 
u 


o 


© 




o 




o 


o 




ON 

rn 




o 




o 




o 




9 e 
Ed A 


■3 
I 

H 


R 

VI 


E 


m 


o 




o 




o 


o 








o 




o 




o 




H A 


c 
« 


% 


E 




o 




o 




o 


>o 




(N 




o 




o 




CT) 




s e 

U. A 


c 

X 


R 

VI 




NO 


o 




© 




o 






r> 




o 




o 




NO 




k 

£ 2 

a a 


c 
I 


VI 


E 


00 


On 




o 




o 


o 




t> 




o 




o 




00 




u 

£ ? 


c 
I 


■ 

VI 


E 


- 






o 

o 




o 
o 


<N 




"3- 




o 

ON 




o 
o 








"u 


E 




































E 


e 




o 


o 




o 




o 


O 




O 




<N 




o 




O 




H 


1 






































M 


■ 














































































■B c 




2 




































1 1 


t/5 

M 
u 


£ 

SI 




— 
1 


1 




C-J 




NO 


i 




m 




o 




>n 




CO 




1 


J 
U 


DC 




































L. 








































^ u 


m 


■ 




m 


00 




— 




CN 


ON 




f<i 




r~- 




_ 




NO 






c 


| 


NO 


■* 




r^ 




On 


<N 




o 




T 




l> 




00 




|*5 


X 




mt 


o 




o 




O 


— 




*-" 




O 




o 




o 






1 


i~ 


| 


NO 


o 




cs 




NO 


(N 




o 




* 




On 




rNi 




III 


L. 


m 


-c 


NO 


«N 




o 




NO 


V"> 




r-« 




rj 




«r> 




rr 




< 


* 


c 


O 


o 




— 




O 


O 




o 




o 




o 




o 




M 




a 


































-si 


id 

\ 


u 


f 


OO 


t^- 




On 




<N 


NO 




r«» 




o 




VI 










« 


-c 


NO 


CN| 




00 




V> 


NO 




NO 




(N 




«r> 




en 




US I 


> 
< 


* 


a 


O 


o 




O 




© 


o 




o 




O 




o 




o 












o 

OS 










£ 




-^ 




















Jj 
■ 

55 






CO 


a 

x: 


'3 


t-H 

0> 
V) 
Vi 

u 


'C 
'3 


O 


V> 

6 

> 

CO 
E 


o 

O 

E 
1 


CO 
VI 

B 

E 
O. 




X) 

3 
O 


VI 

(U 
1 


1 

3 


'C 

'3 


O 


•c 

1 

a. 










QQ 


u 


(X 


U 


a, 


U 


U 3C 


u u 


a 


J 


a 


0- 


tu 



269 



T3 
<L> 

3 
C 

I 

5 

u 

I 
I 

m 
■ 

H 



^ "° E 
































o 5 g 
































S ■ a 


O 




o 




o 




o 




O 


o 


o 


o 


o 




o 


m"; 
































A VI 
































fe' E 
































E > " 

E a ° 


O 




o 




o 




o 




o 


o 


o 


o 


o 




o 


tl § n 
































Error 
>50 
and 
<60 
cm 


o 




o 




o 




© 




o 


o 


o 


"tj- 


o 




o 


Error 
>40 
and 
<50 
cm 


o 




o 




t> 




o 






00 


o 


m 


o 




o 


Error 
>30 
and 
<.40 
cm 


© 




o 




fN 




o 




so 




o 


fN 


o 




o 


Error 
>20 
and 
<30 
cm 


o 




o 




so 




s 




fN 
fN 




o 


00 


o 




o 


Error 
>10 
and 
<20 
cm 


o 




o 




fN 




</-> 




so 


-— 


o 


OS 


o 




o 


|*12 E 

i A g VI « 


00 




00 

os 




- 








<n 


- 




- 


o 

o 




o 
o 


"- I 
































I • 


T 




o 




o 




o 




o 


o 


Os 


_ I 


o 




o 


U II 
































-* S 
































v St 
































■b e "5 »r 
































>del-C 

Statio 
eckSt 
ange( 


SO 




I 




r» 




o 

■ 




*-H 


•-h 


V) 


fN 
i 


00 




v-> 


s n « 
































ll 


































00 




o 




o 




o 




<n 


00 


v> 


«o 


tT 




Os 


1 1 & 1 s 


SO 




SO 




«n 




o 




fN 


V) 


so 


so 


SO 




f> 


AC 


O 




o 




~^ 




~~ 




"* 


1— < 


o 


'— 


d 




© 


Best 

Model 

Estimated 

Average 

Water 

Depth (m) 


_ 




fN 




p» 








fN 


OS 


fN 


so 


p» 




r*> 


so 




"3- 




K 




fn 




SO 


SO 


T 


so 


K 




</*> 


o 




o 




o 




o 




O 


o 


o 


o 


o 




o 


heck 

ation 

mated 

erage 

ater 

th(m) 






so 




00 








o 


l> 


OS 


o 


so 






Vt 




# 




SO 




"•* 




SO 


so 


m 


r^ 


T 




■<* 


h ^ a 


o 




o 




o 




o 




O 


o 


o 


o 


o 




o 


c 
o 

■ 


e 
c 

O 




I 

C/l 

c 

I 


0) 
00 


4> 

i 


U 
> 


60 

"o, 


— 

03 


u 

■g 

1 

CU 

Cl 
O- 


oo 


2 


*v3 
CO 

H a 

3 *S 


t/5 

<u 

•3 


E 


1 

e 8 






s 


Pi! 


3 
00 


2 


00 


>- 
CU 


00 


y 

00 


y 

00 


&o O 


o 
oo 


U 















8 % of biweekly intervals (n=336) in 1980-1993 with difference between modeled and recorded/estimated with-sill water depths by 
10 cm increments. 






Error 
>70 and 
<100cm 


O 





Error 
>60and 
<70cm 


O 





Error 

and 
<60 
cm 


O 





Error 
>40 
and 
<50 
cm 


O 





Error 
>30 
and 
<_40 
cm 







Error 
>20 
and 
<30 
cm 


O 


- 


Error 
>10 
and 
<20 
cm 


SO 
CN 





£9 IS E 


>n 





Error* 

--- (1 cm 


O 


m 


Model-Check 

Station/ 

Check Station 

Range (%) 


(N 


On 


Recorder 

Water 

Depth 

Range 

(m) 


00 




00 


Best 

Model 

Estimated 

Average 

Water 

Depth ( m) 


© 


00 




<u 

3 
C 

I 

O 

u 

T 
i 

m 
■ 

Q 

x> 

CO 

H 




Check 
Station 

Estimated 

Average 

Water 

Depth (m) 


CN 




00 

© 


B 
O 

a 


Territory 
Prairie 




vO 

B 

u 

C/2 

c 
2 
H 













271 

Table 3-4. Summary statistics of recorder data and model output at check stations during 
1980-1993. 



Station 




Mean 








(n=336 biweekly 


Condition 


Water 


Standard 


Minimum 


Maximum 


intervals) 




Depth 

(m) 


Deviation 






Billy's Lake 


With sill 


0.66 


0.23 


0.32 


1.64 




No sill 


0.38 


0.22 


0.10 


1.45 




No outflow 


1.19 


0.21 


0.72 


1.83 




Recorder 


0.68 


0.28 


0.10 


1.73 


Chase Prairie 


With sill 


0.20 


0.14 


0.00 


0.63 




No sill 


0.12 


0.13 


0.00 


0.59 




No outflow 


0.21 


0.16 


0.00 


0.66 




Recorder 


0.27 


0.11 


0.00 


0.48 


Chesser Prairie 


With sill 


1.02 


0.23 


0.48 


1.63 




No sill 


0.97 


0.23 


0.44 


1.55 




No outflow 


1.10 


0.24 


0.54 


1.72 




Recorder 


0.89 


0.22 


0.21 


1.36 


Coffee Bay 


With sill 


0.66 


0.19 


0.24 


1.18 




No sill 


0.62 


0.18 


0.22 


1.10 




No outflow 


0.74 


0.20 


0.28 


1.29 




Recorder 


0.52 


0.18 


0.01 


0.92 


Craven's Hammock 


With sill 


0.52 


0.25 


0.10 


1.52 




No sill 


0.32 


0.24 


0.00 


1.37 




No outflow 


0.92 


0.24 


0.42 


1.71 




Recorder 


0.66 


0.26 


0.08 


1.36 


Cypress Creek 


With sill 


0.70 


0.24 


0.20 


1.36 




No sill 


0.26 


0.18 


0.00 


0.84 




No outflow 


0.45 


0.22 


0.00 


1.11 




Recorder 


0.67 


0.23 


0.18 


1.22 


Double Lakes 


With sill 


0.24 


0.19 


0.00 


0.86 




No sill 


0.22 


0.18 


0.00 


0.84 




No outflow 


0.27 


0.20 


0.00 


0.89 




Recorder 


0.20 


0.12 


0.00 


0.47 



Table 3-4-continued. 



272 



Station 




Mean 








(n=336 biweekly 


Condition 


Water 


Standard 


Minimum 


Maximum 


intervals) 




Depth 
(m) 


Deviation 






Durdin Prairie 


With sill 


0.59 


0.19 


0.25 


1.11 




No sill 


0.56 


0.18 


0.24 


1.07 




No outflow 


0.62 


0.20 


0.25 


1.17 




Recorder 


0.55 


0.15 


0.17 


0.89 


Floyd's Prairie 


With sill 


0.42 


0.20 


0.03 


0.90 




No sill 


0.26 


0.20 


0.00 


0.87 




No outflow 


0.63 


0.22 


0.10 


1.15 




Recorder 


0.31 


0.16 


0.00 


0.86 


Gannett Lake 


With sill 


0.61 


0.23 


0.17 


1.20 




No sill 


0.60 


0.22 


0.18 


1.17 




No outflow 


0.66 


0.23 


0.18 


1.23 




Recorder 


0.51 


0.15 


0.16 


0.83 


Honey Prairie 


With sill 


0.54 


0.30 


0.12 


1.60 




No sill 


0.47 


0.30 


0.07 


1.59 




No outflow 


0.52 


0.32 


0.07 


1.65 




Recorder 


0.77 


0.26 


0.20 


1.37 


Kingfisher Landing 


With sill 


0.28 


0.12 


0.05 


0.60 




No sill 


0.23 


0.13 


0.01 


0.59 




No outflow 


0.25 


0.14 


0.01 


0.64 




Recorder 


0.56 


0.16 


0.17 


0.95 


Moonshine Ridge 


With sill 


0.42 


0.21 


0.00 


0.98 




No sill 


0.38 


0.19 


0.00 


0.92 




No outflow 


0.49 


0.22 


0.00 


1.07 




Recorder 


0.46 


0.14 


0.11 


0.80 


Suwannee River 


With sill 


0.77 


0.21 


0.44 


1.52 




No sill 


0.49 


0.22 


0.15 


1.37 




No outflow 


1.71 


0.24 


0.83 


2.26 




Recorder 


0.68 


0.27 


0.10 


1.60 


Sapling Prairie 


With sill 


0.31 


0.19 


0.00 


0.87 




No sill 


0.19 


0.19 


0.00 


0.85 




No outflow 


0.26 


0.21 


0.00 


0.92 




Recorder 


0.41 


0.18 


0.00 


1.00 



Table 3-4-continued. 



273 



Station 




Mean 








(n=336 biweekly 


Condition 


Water 


Standard 


Minimum 


Maximum 


intervals) 




Depth 

(m) 


Deviation 






Sapp Prairie 


With sill 


0.62 


0.26 


0.06 


1.47 




No sill 


0.32 


0.29 


0.00 


1.47 




No outflow 


0.44 


0.30 


0.00 


1.52 




Recorder 


0.60 


0.22 


0.20 


1.45 


SCFSP 


With sill 


0.69 


0.23 


0.33 


1.62 




No sill 


0.38 


0.22 


0.08 


1.41 




No outflow 


1.33 


0.21 


0.82 


1.83 




Recorder 


0.67 


0.27 


0.10 


1.68 


SCPvA 


With sill 


0.42 


0.13 


0.00 


0.82 




No sill 


0.41 


0.13 


0.00 


0.80 




No outflow 


0.45 


0.14 


0.00 


0.87 




Recorder 


0.39 


0.14 


0.08 


0.74 


Seagrove Lake 


With sill 


1.45 


0.24 


0.43 


2.08 




No sill 


1.40 


0.23 


0.43 


2.01 




No outflow 


1.52 


0.25 


0.44 


2.16 




Recorder 


0.37 


0.15 


0.01 


0.70 


Brown Trail (Sill) 


With sill 


0.21 


0.29 


0.00 


1.27 




No sill 


0.12 


0.22 


0.00 


1.27 




No outflow 


1.17 


0.28 


0.40 


1.81 




Recorder 


0.74 


0.30 


0.13 


2.07 


Sill Gate (South) 


With sill 


0.66 


0.25 


0.24 


1.46 




No sill 


0.38 


0.23 


0.07 


1.32 




No outflow 


1.50 


0.25 


0.73 


2.09 




Recorder 


0.70 


0.27 


0.11 


1.77 


Soldier's Camp 


With sill 


0.77 


0.26 


0.14 


1.42 




No sill 


0.73 


0.25 


0.12 


1.36 




No outflow 


0.84 


0.27 


0.17 


1.53 




Recorder 


0.46 


0.14 


0.17 


0.80 


Suwannee Creek 


With sill 


0.06 


0.15 


0.00 


0.88 




No sill 


0.05 


0.14 


0.00 


0.87 




No outflow 


0.06 


0.15 


0.00 


0.93 




Recorder 


0.56 


0.19 


0.00 


0.95 





Table 3-4— continued. 












274 




Station 




Mean 












(n=336 biweekly 


Condition 


Water 


Standard 


Minimum 


Maximum 






intervals) 




Depth 

(m) 


Deviation 








Sweetwater Creek 


With sill 


0.53 


0.18 


0.18 


1.08 






No sill 


0.47 


0.17 


0.14 


0.99 








No outflow 


0.59 


0.19 


0.22 


1.17 






Recorder 


0.41 


0.18 


0.00 


0.79 






Territory Prairie 


With sill 


0.53 


0.24 


0.03 


1.16 








No sill 


0.32 


0.22 


0.00 


0.93 








No outflow 


0.43 


0.25 


0.00 


1.09 








Recorder 


0.52 


0.16 


0.01 


0.88 






Transect 52 


With sill 


0.13 


0.23 


0.00 


1.03 








No sill 


0.05 


0.16 


0.00 


1.03 








No outflow 


1.00 


0.28 


0.27 


1.74 








Recorder 


0.74 


0.29 


0.13 


2.13 






Transect 55 


With sill 


0.44 


0.31 


0.04 


1.48 








No sill 


0.08 


0.20 


0.00 


1.11 








No outflow 


1.44 


0.28 


0.69 


2.10 








Recorder 


0.74 


0.29 


0.13 


2.09 






Transect 60 


With sill 


0.81 


0.21 


0.44 


1.46 








No sill 


0.67 


0.24 


0.32 


1.51 








No outflow 


1.61 


0.25 


0.66 


2.22 








Recorder 


0.68 


0.27 


0.11 


1.59 














■ 















275 
(Figure 3-12). Although the changes in the Suwannee River floodplain are easily 

attributed to the absence of the sill from the topographic surface, the changes at Cypress 
Creek and Sapp Prairie are puzzling. The hydrologic connectivity that exists between 
these stations and the sill area is outside of the refuge perimeter, in the Cypress Creek 
and Suwannee River drainages, and not within the swamp. Water level fluctuations in 
these areas are significantly correlated with those in the sill region during low and 
average water level conditions regardless of the sill's presence, although this relationship 
is weak. During high water conditions when the sill is present, water depths in the 
Cypress Creek basin decline and then increase with depths in the sill gate area, while 
Sapp Prairie water depths remain positively correlated with increasing water depths at 
the sill (Figure 3-13). This means that in high water conditions, levels at Cypress Creek 
decrease while those at the sill and Sapp Prairie are increasing. Drainage from the 
Cypress Creek area may be affected by variations in the hydraulic head at the creek-river 
junction created by the sill's impoundment of the Suwannee River. The hydraulic head 
must shift as more water is impounded at the sill, increasing the creek-river water surface 
elevation difference at the junction as more water flows freely from the creek and 
therefore from Sapp Prairie. As water levels decrease in both areas this difference may 
become smaller, decreasing drainage of the Cypress Creek area (Figure 3-14). The 
hydrologic environment of both areas at high water behaves independent of the sill gate 
area when the sill is removed from the topographic surface. Without the sill in place, the 
hydraulic head between the creek and river may be reduced, creating conditions for 
slower de-watering of the creek basin. More water is retained in the creek and at the 



276 




+O.06 



10 



10 



20 Kilometers 



Sill 

/sy Refuge Boundary 
A/ Islands 
■■ Islands 
/\/ 5 cm Contours 



N 




Figure 3-12. Inverse-distance-weighted, contoured estimates of increases in average 
semi-monthly water surface elevations (m) at recording stations, attributed to the 
Suwannee River sill during 1980-1993. 






277 



39 



-38 
CO 

1 

E 

I 37 

> 
© 

UJ 



8 

t 

w 36 



£ 



35 - 



Cypress Creek 
Sill Gate 
Sapp Prairie 



l ■ : |f 



5» : :s;%: ;,! ; "t I •. u 1 , i i n i 

n* - H.E. . :. • <. i 1 I I lli I 1 I 



f 




T 



T 



T 



T 



Low Average High 

Water Surface Elevation at Sill Gate (m AMSL) 



Figure 3-13. Comparison of semi-monthly water surface elevations at Cypress Creek and 
Sapp Prairie under increasing water level conditions in the sill gate area during 1980- 
1993. 



278 



* 38 

< 
E, 

"c 37 

o 

s 

(0 

> 
© 

111 

© 
o 

I 

3 
CO 

© 

4-* 

(0 

5 



c 
o 

© 

HI 

« 

3 
CO 

© 

4-* 

CD 

5 



36 - 



35 - 



34 



Low With-Sill 
Average With-Sill 
High With-Sill 
Very High With-Sill 



A 




zi 38 - 
s 

< 



37 - 



36 - 



35 - 



34 



Low No-Sill 
Average No-Sill 
High No-Sill 
Very High No-Sill 




Sill River Creek-River Creek Prairie 

Cypress Creek Watershed Location 



Figure 3-14. Comparison of average, semi-monthly water surface elevations in the 
Cypress Creek watershed under low, average, high, and very high water levels in the sill 
gate area during 1980-1993. 



279 
creek-river junction as it leaves the swamp, allowing a backup of water into the creek 
with additional precipitation (Figure 3-15) Sapp Prairie does not show this inverse 
relationship under high water conditions, indicating that this area is not affected by 
accumulating backwater causing the head reversal in the Cypress Creek and river basins. 
The Sweetwater Creek drainage basin is similarly affected by increased water volume 
with sill removal. However, like Sapp Prairie, the Suwannee River backwater effect is 
diluted before it reaches the creek (Figure 3-16). Water levels at these stations may also 
be affected by activities in the adjacent perimeter areas under timber production (such as 
increased surface runoff into the Suwannee River and area creeks due to clear cutting 
and ditching), which may also be impacting the region's drainage patterns independent of 
water levels in the sill impoundment. 

Changes in water depths do not necessarily mean changes in duration of 
inundation (hydroperiod). To determine if water depth increases were accompanied by 
longer periods of inundation, water depths were partitioned into 7 groups (Table 3-5; see 
Chapter 6 for discussion of interval choice), and number of intervals in each depth group 
during 1980-1993 were tallied and compared between "with-sill" and "no-sill" model 
runs with contingency tables (G-statistic). All areas with significant changes in 
hydroperiod group frequencies also had some increase in average water depth with the 
sill in place (Table 3-6). Not all areas with water depth increases also experienced 
significant changes in frequencies of hydroperiod groups, however. Chesser Prairie, 
Coffee Bay, Double Lakes, Durdin Prairie, Gannett Lake, Moonshine Ridge, and SCRA 
areas increased average water depths 0.01-0.05 m without significant changes in 



280 



E 
I 



g 



co ri ro co 



LU 

fco CO o 

ft U) ££ 2 i* &o 



J I 
o 



PJ 




U 

l- 

u 

CO 

SO 



u 

c 



> 



c 
c 

3 



C3 
1> 

c 



.2 

Q. 
T3 

O 

O 
55 

u. 

> 

2 

i> 

<L> 

C 

I 

3 

on 
JS 



00 

o 

Q. 

& 

a 



C/l 

c 
o 

1 

o 

o 



I 



281 



<t 38 " 

I 

< 

E 

T 37 
o 

a 

I 

% 36 



3 
(0 

i 



5 
© 

LU 



o 
(0 

t 

3 
(0 



35 



34 



- Low With-Sill 

- Average With-Sill 

- High With-Sill 

- Very High With-Sill 




si 38 " 

S 
< 

E 

^ 37 
o 



36 - 



35 - 



34 



Low No-Sill 
Average No-Sill 
High No-Sill 
Very High No-Sill 




T 



T 



T 



Sill River Creek-River Creek 

Sweetwater Creek Watershed Location 



Figure 3-16. Comparison of average, semi-monthly water surface elevations in the 
Sweetwater Creek watershed under low, average, high, and very high water levels in the 
sill gate area during 1980-1993. 



282 



Table 3-5. Water depth ranges for hydroperiod group delineations. 



Hydroperiod Group 


Water Depth Range (m) 


1 


< 0.00 m 


2 


0.00 < depth < 0.05 m 


3 


0.05 < depth < 0.15 m 


4 


0.15 < depth < 0.30 m 


5 


0.30 < depth < 0.60 m 


6 


0.60 < depth < 1.00 m 


7 


depth > 1.00 m 



283 



3 

E 



•8 
E 



I 

c 

— ~ 

— 0) 
t/5 £ 

•a s 

> ° 

• E «J 

6 on 
'o on 
c -r 

i <=> 

= oc 

IB — 

o..S 

II 

O U 
II 

2 fc 
•o -o 

s i 

■a S 

c •- 
ea n 

jj O 

ca P 

? f2 

•E ^ 

ft? 

1 g 

d_i o. 

6 "5 
o > 

.23 g 

S.-2 

E S 

SO On 

■S ^ 

3 on 

H — 



. ] s a 


















<* 










ilillli 


«1 
















m" 










Mi 

Freq 

Hydro 

Gro 

No 
Stmul 
1980- 


»» 
















<N 

pal 










>st 

uent 

period 

ups 

Sfll 

atlon 

1979 


















■t 










Ml 
Freq 

[ydro 
Gro 
No 

slum 
1970 


















c> 






































Mt 

uent 

period 

ups 

sin 

atlon 
1969 




























WI 
















TT 










M 

Freq 
Hydro 

No 
Sbnu 
1960 


Tt- 
















m 










Most 
Frequent 
Hydro period 
Groups 
WHhSIU 
Simulation 
1950-1959 


















cvf 










Most 
Frequent 
ydro period 
Groups 
With SUI 
•Intulatlon 
941-1949 


















t*f 










S 




























Most 
sequent 
roperlod 
ups. With 
Sfll 


v-T 
















l/N 










Ell 


















O 










s a 




























Most 

Frequent 

Hydroperlod 

Groups, No 

SUI 


VI 
















Cvf 










t "t 1 a? 

= lllif 


s 




§ 


>v 




o 

e 






p 


o 

c 


o 

c 


8 
>> 


o 


f § £W 




























U 




























J ■ 


«*> 










O) 


NO 




(S 


s 








1^ 


o 




p" i 


o 




o 


© 






© 


o 


© 


St 


o 




o 


© 




o 


© 




© 


© 


© 


© 


o 


55 a 




























lltlfilf 


\o 
©' 

1 




CO 

O 

©' 

1 


© 




NO 

o 
©' 


o 




VO 
r\| 

©' 


© 

©' 
• 


© 
© 


© 
© 


00 

© 
©' 


li 


ON 




o\ 


Ov 




o\ 


o 




ON 


Ov 


Ov 


Ov 




•<r 




vi 


vO 




t-. 


ON 




•» 


>r> 


VO 


r^ 


Ov 


Ov 




ON 


ON 




ov 


On 




Ov 


Ov 


ON 


ON 


Ov 


"3 ± 


" " 


vO 


T o 


i 


o 


"7 o 




VO 


"7 vo 


"7 © 


J © 


J © 


"7 vo 


^ 


"" 


o •» 


o 


•* 


© Tf 


© 


CI 




© TT 


o t 


© "S" 




J* 




>n cm 


NO 


CN 


r- <n 


00 


CI 


•«»■ <N 


■^1 <N 


VO <N 


r~- cm 




on 


ii 


Os II 


ON 


II 


ON II 


On 


II 


ON II 


Ov II 


Ov II 


Ov II 


Ov II 




"~ 


e 


— C 


^ — 


c 


— c 


"^ 


c 


— e 


— i C 


— c 


— fa 


—" c 


. 


»i 
















o H 










1 


£ 


o 

1 














■c s 












S3 


-J 














£ 



































284 




uent 

period 

ups 

Sill 

ation 

1993 


t-- 










so 












sf£Eo1s 


so" 










M-r 












X ■ 






















Most 

Frequent 

Hydroperiod 

Groups 

No SIM 

Simulation 

1970-1979 


r- 










t- 

so* 










_ a e o> 
























Most 
Frequeni 
Hydro peri 
Groups 

Nil Sill 

Simulatlo 
1960-196 
























r»- 










so 












Most 

Frequent 

Hydroperiod 

Groups 

With Sill 

Simulation 

1950-1959 


f- 










SO 










ml - S* 
























ilillil 


r^ 










SO 

w-T 










.11 
























Bt? 
























Most 
Freque 
ydrope 
roups, \ 

sm 


r~ 

SO* 










so 
iff 












a o 






















-I? 1 
























1 * lie 


r^. 










SO 
























wf 












S J 
























i iifif 
























S £ IJ £§ 
Jfllli 


o 

c 


■ 


g 


g 


o 

c 


o 

c 


en 


o 

c 


o 

c 


o 

c 




Ss^ 
























■ 
























IJ 


rsl 


VO 


© 






rsl 




© 




rsl 




"H -I 


O 


<N 


© 


o 


© 


© 


~- 


© 


© 


© 




11 


©' 


©' 


© 


©' 


©' 


© 


© 


©' 


© 


© 






o 


V) 


rsl 

o 


rsl 

O 


© 


(s-l 

© 


© 


© 


rsl 

o 


</1 

o 




© 


o 


©' 

1 


© 


©' 


©' 


©' 


©' 
i 


©' 


©' 


D 

3 




IJ 


Os 

ON 


os 
«n 


Os 
SO 
Os 


Os 

r~ 

Os 




Os 
5 


Os 
>/■> 

Os 


OS 
so 

Os 


Os 
OS 


Os 
Os 


g 




■a ± 
ll 


T so 


"7 o 


"7 © 


"T o 


*7 so 


T so 


"7 © 


j © 


"7 © 


"7 so 


*^2 




«— * —« 


© * 


© T 


o ■» 


© r<"> 




© "9- 


© •q- 


© *j- 


© m 


§ 
o 
u 




■V <N 


m-i rsl 


SO rsl 


t-~ rsi 


oe ro 


«*• rsl 


ui rsl 


so rsi 


t-- rsi 


oo m 




Os II 


Os II 


Os II 


OS II 


Os II 


Os II 


Os || 


ON II 


Os II 


Os || 






— > S 


— c 


— C 


— C 


— e 


— C 


— C 


— e 


— h 


— c 


T 
1 

SO 
1 

m 




| 


11 










S 










1 




1 










1 

o 

































285 



T3 
<U 

3 

9 

■♦-» 
c 
o 
o 

I 
I 

SO 
I 

m 
1 



_ 'e' em 






























Most 

Frequen 

Hydro perl 

Groups 

No SB! 

Simulatlo 

1980-199 


sC 
sri" 


















rr" 










m *S 5 ^ 






























Most 
Frequen 
Hydroper 
Groups 

No SHI 
Simulate 

1970-197 




















SO 










- 'S e »- 






























Most 

Frequen 

Hydroper 

Groups 

v., sill 

Simulate 

1960-196 


mi 




























ost 
uent 
period 
Kips 

latlon 
-1959 




















t 










>/-> 


















m* 










*zis*i* 


ff 


















<s 










ll «a 11 

itltSn 

irfllis 




















«>■- 










„*§ 


sC 




























5l< 




















SO 










ill tl 


•>* 


















V-T 










co* 


















TJ- 










Most 
vquent 
ro period 
raps. No 
SOI 


«r> 


















f 










»" 




























£?.<■ 


m 





























x * 






























if fill 


I 




1 






e 




I 


V? 

O 




I 


g 




c 


B 


1 


*!"" 






























-1 


•t 




so 




© 




en 


v> 




00 


r~ 




© 


v~> 


1 -= 


© 




o 









O 







O 








© 




11 


o 




o 









© 
















© 







o 




©' 




O 

O 
1 





■ 


CM 

O 




00 
O 

O 
■ 


m 





© 
©' 


© 
© 
© 


t> 

©' 


11 


os 




Os 




OS 




Os 


r»> 




Os 


OS 


Os 


Os 




■t 

OS 




OS 




% 




Is 

Os 


g 




9 
Os 


Os 


SO 
OS 


r- 

OS 


OS 

Os 


1! 


1 


SO 


"7 


O 


1 


O 


j © 




SO 


"7 so 


T O 


"7 © 


"7 © 


J SO 


^* 


~^ 


O 


TJ- 


O 


*r 


O TJ- 





O 


^~ — 


O # 


© T 


© * 


© m 


"3- 


si 


«/-> 


<N 


SO 


<N 


IS <s| 


00 


m 


J» fN 


sn rs| 


SO CN 


t-~ ts 


00 m 


Os 


II 


Os 


II 


Os 


II 


Os II 


OS 


II 


Os II 


OS II 


Os II 


OS || 


Os II 




"* 


B 




e 


^ — 


e 


— C 


** 


c 


— c 


— c 


— C 


— c 


— e 






M 


























1 


(A 


X 
















to 










I 


■ 

■ 






























u ac 




























286 





Most 

Frequent 

Hydroperiod 

Groups 

No SID 

Simulation 

1980-1993 


ft 












in 


• 












Most 

Frequent 

Hydroperiod 

Groups 

No SID 

Simulation 

1970-1979 


</^ 












NO 














1- 












<r\ 














Most 

rdro period 
Groups 
No Sill 
i n hi lafion 
960-1969 














NO 














X w 




























Most 
Frequent 
Hydroperiod 
Groups 
With sid 
Simulation 
1950-1959 


«"» 






































NO 














_ 1 - £ * 




























ililsls 


«/-i 












NO 














M 

Freq 

Hydro 

Gro 

WHI 

Shnul 

1941 


T 












V! 














11 




























Most 

Frequenl 

ydroperl 

roups, W 

SOI 


Vl 


























»* 












NO 

w-T 














ft 


























BO 




























Most 

Frequent 

Hydroperiod 

Groups, No 

sm 














NO 














.i.tif 

Hiiii 

fa ll|6 


1 


o 


o 


o 


o 




o 




o 


o 




O 




>. 


e 


1 


c 


c 




c 




c 


c 


c 




*|** 




























T B 




























* * 


«/"> 


* 


o 


<N 


<N 




(S 




(N 


o 


m 


P4 




II 


o 


o 


o 


o 


o 




o 




o 


o 


o 


© 




o 


© 


o 


o 


o 




o 




d 


o 


o 


o 




Mean 
Water 
Depth 
Change 
(With 

sIB 
No SIM) 

On) 


d 

■ 


o 
d 


§ 

d 


o 
© 


o 
o 




o 
© 




o 
o 


© 
© 


00 

o 

d 


o 
6 


-d 




11 


On 


Os 


On 


On 


ro 




On 




On 


ON 


On 




i> 




;» 


«r> 


vO 


r^ 


On 




TT 




V> 


NO 


r» 


On 


3 




On 


on 


On 


On 


On 




ON 




ON 


On 


On 


On 


C 




it 


"T NO 


"7 o 


V O 


T © 




NO 




NO 


T o 


"7 o 


"T O 


"7 NO 


: ! 




^m ~^ 


O TJ" 


© ■* 


o * 


O 


CI 






O rf 


© rr 


O T* 


O ro 


c 
o 
a 




■* (N 


</-> rs 


NO <N 


t^ <N 


oo 


N-O 


« 


^J 


•Ai r-) 


NO <N 


r-~ cs 






ON II 


On II 


On II 


On II 


On 


II 


ON 


II 


On II 


On II 


On II 


On II 






— c 


— C 


— C 


— C 


" "* 


c 


*""" 


c 


— e 


— e 


— C 


— C 


■ 
n© 




























en 
u 




1 


u 

■2 «> 












{ 


O 
















H 












& 


£ 











287 







SO 



■ 1 II 

ntr 



x OZ 



|£ 






o 



illiils 
J! 



J. 



_ 13 ■ 0\ 

llilili 
"lilfll 



•"if" 



llilili 

■fifth 



in 



M 11* 

■lit 

Ell 



vD 
•rf 



SO 



E 1 z 







p 



c 
c 



o 
© 



o 



o 
©' 



o 

o 

©' 



00 

o 
ci 



o 



r- 

o 
©' 



o 
o 

© 



o 
© 
© 



© 
©' 



■u 

3 

c 
"£ 

O 
o 

I 
I 

v© 
i 

— 
CO 

H 






ill 



os 
© 

© 



<N 

© 



SO 

o 



© 

© 



© 
©' 



© 
o 



© 
©' 



o 
© 



© 
©' 



II 



os 

OS 



as 
os 



as 

SO 

OS 



I! 



■ VO 


©°- 


, © 
© tj- 


Tf (N 


•O (N 


SO <N 


OS II 


OS II 


Os II 


— C 


— C 


— C 



OS 

t^ 
Os 

"7 © 

© T 

Os || 



Os 

OS 

"T so 

© o 

oe m 

OS || 



Os 

Os 






OS 
«/"> 

Os 

"7 © 
© ■* 

</-> (N 

OS II 

— < C 



Os 
SC 

Os 

V © 

© T 
SO <N 
Os II 



Os 

P- 
Os 

"7 © 

© Tf 

OS II 



CI 

OS 

OS 

"7 so 

O c> 
00 m 

Os || 

— a 



Si 

u. a. 



288 



-a 

C 



vo 

I 

tn 



_ 1 em 
































l £1 ||ll 

S 3 S. 3 7 « •- 


vo 


















VO 












Mi 

Freq 

Hydro 

Gro 

No 

Slmul 

1980 


V 


















>n 












- - B« 
































S t t O.S - » 

"t 1 ■■ 


VO 


















vO 












_ ^ C ON 
































S 8 £. 3 « « — 


VO 


















VO 












M 

Freq 

Hydro 

Gro 

No 

1960 


V~l 


















«n 












will 

s 


Ml 


















v© 












Hi 

period 

ation 
1949 


VO 


















ve 














w-i 


















w> 












Most 

Frequent 

ydroperlod 

roups. With 

SID 


VO 
irT 


















VD 












so 
































11 ° 
































l«I 


vO 






























Most 

Freque 

llydrope 

Groups, 

SUI 




















VC 












ll 

ill lit 

if fill 
































o 

c 




o 

e 




o 
e 




a 


o 

c 




Vt 

u 
>v 




B 


V) 


1 


I 


*■" 
































■E § 
































3 m 


CM 




o 




o 




^H 


fN 




v> 




w-i 


wm 


CM 


vO 


11 


o 




~* 




© 




o 


© 




© 




© 


© 


© 


© 


II 


o 




© 




© 




© 


© 




© 




© 


© 


© 


© 


lllliitfl 


o 
©' 




s 

©' 




r-i 
© 
©' 




© 

©' 
■ 


vv 

© 
©' 




© 
©' 




VO 
© 
© 


©' 


© 
© 


Ov 
©' 


11 


o\ 




ov 




Ov 




OV 


CO 




Ov 




Ov 


Ov 


Ov 




■*f 




1/^ 




vO 




P~ 


Ov 




•■*■ 




«-> 


vO 


r~ 


Ov 


ov 




Ov 




Ov 




Ov 


Ov 




Ov 




OV 


Ov 


Ov 


Ov 


■a j 

il 


1 


VO 


i 


© 


i 


© 


"T © 


■"T 


ve 




VO 


T © 


"7 © 


i © 


J VC 


^~ 


^ 


o 


•*f 


© 


■O- 


© i- 


© 


m 


^ 


»— 


© •q- 


© •* 


© ■* 




T 


rM 


•o 


(N 


vO 


<N 


t^ CM 


00 


o 


■* 


IN 


W1 CM 


VO CM 


r~ <N 




Ov 


II 


Ov 


II 


OV 


II 


Ov II 


Ov 


II 


Ov 


II 


Ov II 


Ov II 


Ov II 


Ov || 






e 


^™ 


C 


wmi 


a 


— e 


*~ H 


e 


~" 


c 


— C 


— c 


"■ c 


— e 




8 






























e 


§ 
2 


u 

00 

K 
















t/3 


I 











289 



-a 

3 

C 

c 
o 
o 

I 
I 

o 

I 

u 
X) 

CO 

H 



_ 1 B «> 




























Most 
Frequen 
lydroper* 
Groups 
No Sill 
SImulatIo 
1980-199 
























































Most 
Frequent 
[ydroperlod 
Groups 
No SUI 
Simulation 
1970-1979 


















r» 
so" 






































m 1 Bo. 




























Most 
Frequen 
Hydro per 
Groups 

No sin 
SlmulaHo 
1960-196 


















■*•" 










_ 'S E Os 


















a* 










Most 
Frequent 
Hydro perl 
Groups 
With SOI 
SImulatIo 
1950-195 


■/■> 
















ro 
rsf 










Most 

Frequent 

Hydroperiod 

Groups 

WHhSUI 

Simulation 

1941-1949 


«o 
















•* 










Tf 
















rM 










ill 


SO 
















r-* 










Mosl 

Frequc 

yd rope 

roups, 

SUI 


vT 
















w-T 










Tf 
















■«r 










x o 




























! = *.£.= 


w> 
















<o 
































































ii 




























llllll 

SiHIi 


8 




a 


g 


B 




9? 

O 




B 

>> 


t 


B 


B 


p 


■ - 


CM 




SO 


<N 


r^ 




r~ 




«i 


•9 




<N 


«N 


■g 3 


O 




o 


© 


o 




o 




— 


CM 


© 






II 


o 




o 


© 


o 




© 




© 


© 


© 


© 


© 


Mean 
Water 
Depth 
Chanfe 
(WUh 

No SUI) 
(m) 


© 

© 

o 




o 
o 

1 


© 
© 


r~ 

© 




©' 




<N 

O 
■ 


o 


© 
© 


OS 

rs 

© 


© 
©' 


1$. 


©s 




os 


OS 


Os 




O 




Os 


OS 


Os 


Os 


o 


*1- 




«/-> 


SO 


r~ 




Os 




Tf 


"/■> 


SO 


r~ 


Os 


Os 




Os 


Os 


Os 




OS 




Os 


Os 


Os 


Os 


Os 


u 


*~" 


so 


"7 o 


"7 © 




© 




sO 


"7 so 


■ o 


"7 © 


"7 © 


T so 

© m 


^* 




o •* 


© * 


© 


"3- 


© 


ro 




© ■* 


© •* 


© * 


Tr 


~-\ 


«1 <N 


SO CN 


t~ 


r-M 


00 


r*i 


3t <N 


•/^ rsi 


so C-J 

d\ II 


t-~ rM 




Os 


II 


OS II 


OS II 


Os 


II 


Os 


II 


Os II 


Os II 


Os II 


Os II 




"" 


c 


— c 


— c 


•"* 


S 


w * 


c 


— C 


— e 


— c 


— < e 


— c 




00 


























■8 


1 
















*> 










i 


"o. 


L 














S-<i 














t- 














OS £ 












J-. 


x 














co a. 














290 



iliiiii 

I ITIjIi 



iliHii 






lazjS; 



I! 



VO 



• 1 



= i. = 
§■££ = = 



II 






VO 

</-r 



I « a ■ J I 
Iliiiii 



_ S _ c c- 

a If |I|S 

*JFf*«H 




VO 

v-T 



1 lit I 






Z fjj 



.Iilii 

mm I? 






o 

c 



c 
B 



o 



o 
d 



ov 
o 



o 
o 



o 



o 



o 

o 



© 



© 

o 



it ill* l* 



o 



O 

o 



vo 

o 



o 

o 



o 



o 



O 



o 
o 



o 
o 



4> 



i 

en 

x> 
H 



It 
II 



ov 

ov 

*7 vo 

ON II 

— c 



a. 

y 

CO 



ON 

in 

OV 

"7 o 

o * 

OV || 



Ov 
VO 

Ov 

T o 
o rr 

VO <N 
Ov II 

— c 



Ov 

r- 

Ov 

T o 
o t 

OV II 

— e 



Ov 
Ov 

"7 vo 

o m 

O0 ON 

Ov II 

— . c 



OV 

» 

Ov 



VO 

<N 

II 



Ov 
OV 

"7 o 

O T 

Wl <N 

OV II 

— < c 



Ov 
VC 
Ov 

"7 © 

o ■* 

VO (N 
Ov || 



OV 
Ov 



o 
* 

<N 

II 



Ov 

Ov 

"7 vo 

o fl 

oo <"■> 

ov ll 

— < a 



O 

CO 



291 



T3 

<D 

§ 

■ — 

1 
O 

o 

SO 
t 

m 

1 



.. 'S s ■ 


























iHiili 
















W-l 




































Most 
Frequent 
ydro period 

Groups 

NoSUI 
simulation 
1970-1979 


vo 














vO 










X 


























_ "S e on 


























S 3 ft 1 3! m m 
















irf 










nihil 
















vr> 










;>t 8.7! f 9v 

1 1 L f > ] •> 
















IT! 










.ill 
















r-- 










1 S irfs 
















vo" 










el! 
















to 










so 


























Most 

Frequent 

Hydroperiod 

Groups, No 

sin 


vo 






































VO 

•rf 










zl.££«-,~ 


























if llll 


























p 




B 


t 


O 

c 






8 
>* 


8 


O 

c 




c 


I 


*i t,B 


























11 


S 




Cvl 




•<*• 


ON 




rs 


«-» 




* 


rs 


"S J! 




O 








O 




*■( 











O 


Ik 







O 








O 







© 





© 


© 




r- 




r~ 

O 


00 




vo 
© 


O0 




Ov 


O 


© 


Ov 
© 


g 


ZZ&£t*i" 


O 




O 
















O 





©' 





11 


Ov 




ON 


C\ 


Ov 


m 




OV 


Ov 


Ov 


Ov 




•*• 
Ov 






VO 
Ov 


t-~ 

Ov 


g 




1- 

o\ 


vn 

Ov 


VO 

O 


r^ 

Ov 


£ 


J J 


1 


VO 


"7 


"7 


"V 


^ 


vO 


T vo 


"T 


7 © 


j © 


"T vo 


** 


^ 


■* 


•<»• 


O T 





r'l 


. 1 — 


O T 


© TT 


© f 




•<J- 


<N 


V) (N 


vO <N 


r-- <n 


00 


Cl 


■» <N 


u-i rM 


vo r<i 


r~ rs 




ov 


II 


OV II 


OS II 


ov II 


ON 


II 


ON II 


Ov || 


Ov II 


Ov II 


Ov II 




— 


C 


— c 


— C 


— B 


~ — 


c 


— C 


— C 


— c 


— C 


— B 




- 














i 










1 


a 


5 
3 













h 












e/5 


^ 












in U 



































292 




* 1 59 


























g = ft 1 I 1 - 


V~ t 










SO 














Freq 

Hydro 

Gro 

No 
Slmul 
1980 


V 










w-r 












— ^ E> 
























Most 
Frcquen 
Hydroperi 
Groups 

N.i Sill 
Kbnulatlu 
1970-197 


</-> 










r-» 














w 










SO- 












Most 

Frequent 

Hyd roperlod 

Groups 

NoSIO 

Simulation 

1960-1969 












SO 










* £> SB - 












SO 

vs" 










_ 1 _ e a 

1 7 £ £ 1 1 - 
£ *- 


VI 










so 










- 1 5- * = 

s J J I 7 


SO 
Iff 










SO 












fit J 


f' 






















■ I* 


v> 










SO 












s fl ?* 


V 










vs." 














B fj 
























■ r 1 1 e 9 

■1 M Jf 

1*1116 





B 


o 

c 


. 


B 


1 


B 


8 

>. 


B 


8 

>*s 




■ 
























TJ 


§ 


<N 


Q 


© 


Os 


Cr> 


o 




so 


SO 




"2 I 


© 


o 


o 


o 


o 


© 


© 


© 


© 






II 


o 


©' 


C> 


d 


© 


©' 


© 


©' 


©' 


©' 






Mean 
Water 
Depth 
Change 
(WUh 

sin- 
No Sill) 

(m) 


v> 


o 


© 


o 


«N 


vs. 


Os 
© 


vs. 

© 


rs 


** 






o 


©' 


©' 


© 
■ 


O 


© 


©' 


© 


©' 
i 


©' 




-d 




1* 


CT> 


o\ 


OS 


Os 


f> 


OS 


Os 


Os 


OS 


ro 


6 




* 


vi 


SO 


r~ 


OS 


Jf 


v> 


SO 


r- 


Os 




3 




* 


ON 


OS 


os 


Os 


OS 


OS 


Os 


Os 


Os 




C 




It 

f] 


T so 


"7 o 


T o 


"7 o 


"7 so 


T so 


j © 


"7 © 


"7 © 


"7 SO 




*£i 




— »■* 


o •* 


o ^r 


© •» 


© m 




© ■* 


© ■t 


O f 


© Ci 




c 




*r cm 


VS «N 


SO Csl 


t-~ rsi 


oo m 


■* r-j 


v> <N 


so rsi 


r~ r-i 


00 o 






OS II 


OS II 


Os II 


OS || 


Os II 


Os II 


Os II 


Os II 


OS || 


OS || 




o 






— C 


— c 


— i c 


— e 


— c 


— c 


— C 


— B 


— C 


— c 




o 


















































o. 




m 


fc> 










M 












a 




* 


O o 










o 












3 






U 










H so 









































293 
hydroperiod group frequencies (Table 3-6). There were also areas with as little as 0.06 m 
increase in average water depth that had significant changes in hydroperiod group 
frequencies (Figure 3-17). Whether these changes are significant to the swamp 
vegetation composition most likely depends on the timing and type of hydroperiod 
changes occurring (See Chapters 6 and 7). 

Changes in frequencies of water depths measured in the 7 "hydroperiod groups" 
with and without the sill present during 1941-1993 generally were experienced in the 
Suwannee River drainage area in the western half of the swamp, but not in the eastern 
swamp (Figure 3-18). The areas surrounding recorders in Billy's Lake, SCFSP, Floyd's 
Prairie, Sapling Prairie, Craven's Hammock, Suwannee River, Sill Gate, Transect 60, 
Cypress Creek, Sapp Prairie, and Chase Prairie had longer periods of slightly deeper 
water depths during 1960-1993 (with-sill) than during 1941-1959 (no-sill) (Table 3-6), 
with greatest duration of flooding occurring during 1970-1979. All areas experienced 
elevated water levels and prolonged hydroperiods during 1970-1979 regardless of the sill 
condition (sill present or absent in model simulations). Similarly, during 1980-1993 and 
1950-1959 water levels were comparatively lower in all parts of the swamp regardless of 
the sill's presence in the model simulation (Table 3-6). Although inundation duration 
was also slightly greater during 1970-1979 than other intervals at Sweetwater Creek, 
inundation depths and frequencies in the remaining areas (Coffee Bay, Sweetwater 
Creek, SCRA, Chesser Prairie, Gannett Lake, Durdin Prairie, Double Lakes, Moonshine 
Ridge, and Territory Prairie) did not change during 1941-1993 with the addition of the 
sill during 1960-1993 (Table 3-6). During the simulated "no-sill" condition of 1960- 



294 




4,5,6 to 3,4,5* 
(-0.44) 






Figure 3-17. Changes in most frequent hydroperiod groups during 1980-1993, with sill 
removal. Numbers represent the most frequent with-sill hydroperiod groups (see Tables 
3-5 and 3-6) versus most frequent no-sill hydroperiod groups. Areas with significant 
change are marked with *. Average semi-monthly water depth decrease with sill 
removal is noted in ( ). 



100 



80 - 
60 
40 
20 



100 



80 - 

60 - 

40- 

20 


'100 



« 80 - 
I 60- 

! — 

o 

1 20- 

B, 
o 



100 



80 - 
60 - 
40- 
20 - 




100 



80 
60 - 
40 - 
20 




Recorder 
■ No Sill 

With-Sill 



T 



J Recorder 
3 No-Sill 
"™~] With-Sill 



J Recorder 
I No-Sill 
~ With-Sill 



H 



Recorder 

No-Sill 

With-Sill 



Recorder 

No-Sill 

With-Sill 



Billy's Lake 



295 



1941-1949 



JZL 






■■ □_ 





1 *—-!_ 




1980-1993 




<=0 .001 -.05 .051 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-18. Changes in hydroperiod depth group frequencies with and without the sill 
during 1941-1993, by decade intervals. 



Chase Prairie 



296 




<=0 .001-.05 .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 

Hydroperiod Depth Group (m) 



Figure 3- 18-continued. 



100 



80 - 
60 - 
40 
20 - 




100 



80 - 
60 - 
40 
20 - 




"100 



80 - 

1 60- 

o 

1 20 

a 
p 



100 



80 
60 - 
40 
20 




100 



80 - 
60 - 
40 - 
20 - 




J Recorder 

r 7 ! no sili 

1 With-Sill 



J Recorder 

| No-Sill 
With-Sill 



J Recorder 
2 No-Sill 
1 With-Sill 



J Recorder 
I No-Sill 



With-Sill 






Recorder 

No-Sill 

With-Sill 



Chesser Prairie 



1941-1949 



JZL 



T 



1950-1959 



n 




1960-1969 




1970-1979 



1980-1993 



I r^*** 



I I I T 
<=0 .001 -.05 .051 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 

Hydroperiod Depth Group (m) 



297 





Figure 3- 1 8-continued. 



100 



80 - 

60 - 

40- 

20 


100 



80 - 

60 

40- 

20 - 


'100 



5 80 
2 

c 



60 - 



o 

§ 

I 

o 

a 
o 



40 - 



20 



100 



80 - 
60 
40 - 
20 - 




100 



80 - 
60 - 
40 - 
20- 




J Recorder 
~ No Sill 
With-Sill 



J Recorder 
~~ No-Sill 
With-Sill 



J Recorder 
1 No-Sill 
1 With-Sill 



Recorder 



□ No-Sill 
1 With-Sill 



j Recorder 
1 No-Sill 
With-Sill 



Coffee Bay 



298 



1941-1949 



n 



1950-1959 



1960-1969 




1970-1979 




1980-1993 



J 1^,,-hlHM 




■A^-l 



<=0 .001 -.05 .051 -.15 .151.30 .301 -.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-18~continued. 



100 



80 - 

60 - 

40- 

20- 


100 



80 - 
60 - 
40 
20 - 



£100 

■ 
« 80 

f 60- 

140- 
o 

% 20- 

I o 

100 



80 - 
60 
40 - 
20 - 




100 



80 - 
60 - 
40 
20 - 




Craven's Hammock 



299 



J Recorder 

J No snl 

1 With-Sill 



1941-1949 



np 




T 



J Recorder 
I No-Sill 
1 With-Sill 



JZL 




n 



J Recorder 

( 1 N °- Si » 
I With-Sill 



n 



XX 



u. 



:■■:';■ 



L_ 





Recorder 

No-Sill 

With-Sill 


1 





XI 



J Recorder 
J No-Sill 
I With-Sill 



i MM 




1950-1959 



1960-1969 



1970-1979 




1980-1993 



yj 



ru*» 



<-o 



.001-05 .051-.15 .151-.30 .301-.60 .601-1.0 
Hydroperiod Depth Group (m) 



>1.00 



Figurg3-18-gpntinued. 



100 



80 - 

60 - 

40- 

20 


100 



80 
60- 
40 - 
20 - 




£100 

| 80 
| 60 

!- 

o 

1 20 

o. 

p 



100 



80 - 
60- 
40 - 
20- 




100 



80- 
60 - 
40 - 
20- 




Cypress Creek 




r 



r 



Recorder 
No-Sill 

O With-Sill 



jj 



J Recorder 
I No-Sill 

§ With-Sill 



n 



J Recorder 

§ No-Sill 
§ With-Sill 



J , mi 



B 



Recorder 

No-Sill 

With-Sill 



<=0 



■CZ-L. 




300 



1941-1949 



1950-1959 



1960-1969 




1980-1993 



J=L 



T 



.001-.05 .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-1 8-continued. 



100 



c 

m 
a I 

c < 

1-1 

^20- 

& 
o 



100 



80- 
60 - 
40 
20 - 




100 



80 - 
60 
40 - 
20 




Double Lakes 



301 




J Recorder 
I No-Sill 
" With-Sill 



1960-1969 




J Recorder 

r~~i No-sm 

1 With-Sill 




Recorder 

No-Sill 

With-Sill 



1980-1993 




v?ytt&?A 



I 
<=0 .001-.05 .0S1-.15 .151-.30 .301-.60 .601-1.0 >1.00 

Hydroperiod Depth Group (m) 



Figure 3- 1 8 -continued. 



100 



80 - 
60 
40 - 
20 - 




100 



80 - 
60 - 
40 - 
20- 




S?100 



« 80- 



60 - 



I 40 
.0 

1 20 

a 

I o 

100 



80 - 
60- 
40 
20 




100 



80 - 
60 - 
40 
20- 




Recorder 
No Sill 

With-Sill 



J Recorder 
I No-Sill 
1 With-Sill 



J Recorder 
I No-Sill 
1 With-Sill 



J Recorder 
r~l No-Sill 
| With-Sill 



J Recorder 
I 1 No-Sill 
§ With-Sill 



Durdin Prairie 



302 



1941-1949 



1950-1959 




1960-1969 




1970-1979 






1980-1993 




1 I 1 1 1 1 — 

<=0 .001-.05 .051-.1S .151-.30 .301-.60 .601-1.0 >1.00 

Hydroperiod Depth Group (m) 



Figure 3- 1 8-continued. 



100 



80 - 

60 

40 

20- 


100 



80 - 
60 - 
40 - 
20 



100 
80 - 
60 - 
40 - 

I 

20- 


100 



80 
60 - 
40 - 
20- 




100 



80 - 
60 - 
40- 
20- 




Floyd's Prairie 



J Recorder 
1 No Sill 
With-Sill 



T 



J Recorder 
I No-Sill 
1 With-Sill 



a 



r_i 



J Recorder 
[] No-Sill 



_L-JisSh_ 



T 






J Recorder 
2 No-Sill 
1 With-Sill 




Recorder 

No-Sill 

With-Sill 



nm czEL 




303 



1941-1949 



1950-1959 



1960-1969 




1970-1979 



1980-1993 



<=0 .001 -.05 .051 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-1 8-continued. 



100 



80 

60 - 

40 - 

20- 


100 



80 - 
60 - 
40 
20 




£100 

5 80 

s 

c 



60 - 



o 



o 

a 
o 



20- 



100 



80 - 
60 - 
40 
20 - 




100 



80 
60- 
40 
20 - 




Gannett Lake 



304 



Recorder 
No Sill 

With-Sill 



J Recorder 
| No-Sill 
§ With-Sill 



J Recorder 
| No-Sill 
I With-Sill 






Recorder 

No-Sill 

With-Sill 



g 



Recorder 

No-Sill 

With-Sill 



XI 




1941-1949 



1950-1959 



T-— ' 




1970-1979 



1980-1993 



<=0 .001-.05 .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-1 8 ~continued 



Sapling Prairie 



305 




<=0 .001 -.05 .051 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 
Hydropenod Depth Group (m) 



Figure 3-1 8-continued 



100 



80 - 

60 - 

40- 

20 - 


100 



80 - 
60 - 
40 - 
20- 




S?100 



« 80 

a 



60- 



o 
I 20 

100 



80 
60- 
40 
20 




100 



80- 
60 - 
40 
20 - 




Recorder 
No Sill 

With-Sill 



Recorder 
■ No-Sill 
With-Sill 



J Recorder 

13 No-Sill 
1 With-Sill 



Recorder 

No-Sill 

With-Sill 






. Hi . 






Recorder 

No-Sill 

With-Sill 



Sapp Prairie 



306 



1941-1949 




J=L 



1950-1959 



_Eb_ 



-rp 



1960-1969 



i b^ 



1970-1979 



ra_ 



■ — Wna 




1980-1993 




<=0 .001-.0S .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydropenod Depth Group (m) 



Figure 3- 1 8-continued. 



100 
80 - 
60 - 
40 - 
20 




100 



80 - 
60 
40 - 
20 




£100 



« 80 - 

| 60- 

1 40- 

I 20- 

Q. 
O 



100 
80 - 
60 
40 - 
20 - 



100 
80 - 
60 - 
40 

20 H 



Recorder 
| No Sill 

With-Sill 



J Recorder 
I No-Sill 
§ With-Sill 



J Recorder 
| No-Sill 
I With-Sill 



j Recorder 
| No-Sill 
I With-Sill 



SCFSP 



307 



1941-1949 




XI 



1950-1959 



.*=*, 0= 




ZDm 



1960-1969 



n 




n^ 



1970-1979 



r~-T^U 




Recorder 

No-Sill 

With-Sill 



1980-1993 



<=0 



.001-.05 .0S1-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-1 8 -continued 



100 
80 - 
60 - 
40 
20 - 




100 



80 - 
60- 
40 - 
20 




£100 
» 

80 

1 60 

!«. 

o 

% 20 

a 
o 



100 



80 - 
60 
40 - 
20 - 




100 



80 - 
60 
40- 
20 




SCRA 



J Recorder 
| No Sill 

§ With-Sill 




Recorder 

No-Sill 

With-Sill 



» a. 



J Recorder 
~ No-Sill 
With-Sill 



J Recorder 

3 No-Sill 
1 With-Sill 



J Recorder 
I No-Sill 
1 With-Sill 



I~L 





n 



n 



n 



u. 



308 



1941-1949 



1950-1959 



1960-1969 



1970-1979 



1980-1993 



I 

<=0 .001 -.05 .061 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 

Hydroperiod Depth Group (m) 



Figure 3- 1 8-continued. 



100 



80 - 

60 - 

40 

20 - 


100 



80 

60 

40 

20 


£100 



80 - 
60 - 
40 - 
20 



100 



80 - 
60 - 
40 - 
20 




100 



80- 
60- 
40 - 
20- 





Sill Area (South Gate) 



309 



J Recorder 
| No Sill 

1 With-Sill 



T 



3 Recorder 
| No-Sill 
1 With-Sill 



Recorder 

No-Sill 

With-Sill 



Recorder 
I No-Sill 
With-Sill 



3 Recorder 
~ No-Sill 
With-Sill 



1941-1949 



T 



JZL 



1950-1959 




WL 



_ 



1960-1969 



n 



1970-1979 




1980-1993 



■ ■CSvl 




I I I 

<=0 .001-.05 .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3-1 8 -continued. 



100 



80 

60 - 

40 - 

20- 


100 



80 
60 
40 - 
20- 




S?100 



m 
a 
I 

£ 

c 

i 

c 
o 

I 

o 

a 
o 



80 - 
60 
40 - 
20 - 




100 



80 - 
60 - 
40 
20 - 



100 
80 - 
60- 
40 
20 



Suwannee River 



310 



_ Recorder 
J No Sill 

With-Sill 



T 



J Recorder 
| No-Sill 
§ With-Sill 



Recorder 

No-Sill 

With-Sill 



Recorder 

No-Sill 

With-Sill 



J Recorder 
| No-Sill 
I With-Sill 



1941-1949 



JZL 



I 






1950-1959 



r - i 



1960-1969 




1980-1993 



—I I I 

<=0 .001-.05 .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3- 18-continued 



100 



80 - 
60 - 
40 - 
20 - 



100 



80 - 
60 - 
40 - 
20- 




£,100 

a 

a 80 

| 60 

| 40 

I 20- 

Q. 
O 



100 



80 - 
60 - 
40 
20 - 




100 



80 - 
60- 
40 - 
20 - 




Sweetwater Creek 



3ll 



J Recorder 
| No Sill 

■ With-Sill 



Recorder 

No-Sill 

With-Sill 



T 



J Recorder 
| No-Sill 
1 With-Sill 



Recorder 



I No-Sill 
§ With-Sill 



Recorder 

No-Sill 

With-Sill 




1941-1949 






1950-1959 



„ 1960-1969 



n 




1980-1993 




T 



<=o 



.001 -.05 .061 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3- 18-continued 



100 



100 



80 - 
60 - 
40 - 
20 - 




100 



80 - 
60 - 
40 
20 




Territory Prairie 



312 




J Recorder 
f No-Sill 
~~\ With-Sill 



y 



Recorder 

No-Sill 

With-Sill 



ns^i 




1970-1979 



1980-1993 



KJiii — 



<=0 .001 -.05 .051 -.15 .151 -.30 .301 -.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3- 18-continued 



100 



80 - 

60 - 

40 - 

20 - 


100 



80 
60 - 
40 - 
20- 




£100 

» 
« 80 

| 60 

»«. 

o 

I 20 

a 
o 

100 



80 
60 - 
40 
20 - 




100 



80 - 
60 - 
40 
20 - 




Recorder 
No Sill 

With-Sill 



Recorder 

No-Sill 

With-Sill 



j Recorder 
~ No-Sill 
With-Sill 



J Recorder 
| No-Sill 
H With-Sill 



J Recorder 
~ No-Sill 
With-Sill 



Transect 60 



313 



1941-1949 



I \VM 




1950-1959 



n 




1960-1969 



1970-1979 



n 



£ 



1980-1993 



l~l 




<=0 .001-.05 .051-.15 .151-.30 .301-.60 .601-1.0 >1.00 
Hydroperiod Depth Group (m) 



Figure 3- 18-continued 



314 
1993 and the "with-sill" condition of 1941-1959, frequencies of water depths in each 
hydroperiod depth group indicated a similar sill impact area. No changes in inundation 
duration, or hydroperiod, were indicated at Sweetwater Creek, Territory Prairie, Coffee 
Bay, SCRA, Chesser Prairie, Gannett Lake, Durdin Prairie, Double Lakes, and 
Moonshine Ridge when the sill was added to the 1941-1959 model runs or removed from 
the 1961-1993 iterations (Table 3-6). Slightly lower water depths and shorter inundation 
intervals occurred in the remainder of the check station areas when the sill was removed 
from the 1960-1993 model iterations; longer flooding periods were recorded in these 
areas when the sill was added to the 1941-1959 simulations (Figure 3-18). 

During 1941-1949 and 1980-1993 the presence of the sill increased hydroperiods 
during both the growing and non-growing seasons in the area encompassed by Craven's 
Hammock, Floyd's Prairie, Lower Territory Prairie and western Chase Prairie, SCFSP, 
and Transect 60 (Figure 3-7). During the 1950-1979 "with-sill" simulation, the area 
affected during the growing and non-growing seasons was primarily in the vicinity of the 
sill and Pocket area; fewest stations were affected during 1950-1959 (Table 3-7). 
Regional Hvdrologic Trends 

Changes in relationships between the water levels in the sill area and throughout 
the swamp occurring with sill removal and at various water level conditions do not 
necessarily result in significant changes in hydroperiods. Comparisons of correlation 
relationships between the sill area and locations throughout the swamp under a range of 
water level conditions with and without the sill present suggested changes not apparent in 



315 



c« 




6 




c 




•o 




u 




— 












C/3 




i 




% 




c 


T3 


•■■■1 


u 


B9 


s 


u 




O 


E 


c 


o 


G 

3 


<u 


5" 


rt 


n 






On 


a<? 


3 


*"H 


O 
feh 


O 
00 


-o 


On 


o 




C 


c 


o 
c 

■a 


<4-» 

c 
<u 

E 


>> 


« 


A 


<D 


a 


i 


o 


ns 


<u 


1 


oi) 


o 


c 


u 




<- 


o 


t/3 


g> 


3 




u. 


c 


G 


o 


> 


e 


— 


-o 


f?i 


c 


i 


CO 


x: 



& 



c 'S 
5 ft 

60 a 
C £ 

i '5 

o z. 

c 

3 .2 

i& 

c 

2 <0 

«+- — 
o • 

S ^~ 
S - 

C8 1/3 

S. c 
£ .2 
P a 
0| 

t^ E 

CO "5 

11 

H E 





£E' 



, a- x - 



IllfflJll 

* IIIJ Hit 



iiiiftfl! 



4) 



8 



o 
B 



o 
B 



>1 



o 

c 



g 
c 






a 
B 



0} 






o 
c 



a 
B 









a 
B 



a 
B 






a 
fl 



§ 



£ 
>> 



§ 



o 
B 



o 
c 



a 
B 



o 
c 



a 
fl 



o 

c 



B 






O 
C 






o 
B 



o 
c 






if I'll 







o 
B 



o 
B 



o 
c 



a 
c 



a 

B 



a 
fl 



a 
B 



c 
B 



o 
B 



a 
B 



4> 



o 
c 



a 
fl 



o 
c 



a 
c 



a 
c 



o 
fl 



o 
B 



R 



a 
B 



g 



a 
B 



a 
c 






o 
fl 



a 
fl 



o 

c 



o 
fl 



o 
c 





g 



g 






o 
fl 



s 



o 

fl 



o 
c 



o 
c 



c 
B 



o 
B 



8 
>> 



R 



o 
fl 



<L> 






1> 



g 



g 



K 









a 
c 



t/1 



g 



g 



03 J 



D 



D a, 



8 .£ 

fl 'a 

£ 



03 

l 

o 
O 



o 



i) 



G X u u 




316 



1 I 






B 



CJ 






0) 



o 
B 






o 
B 






1 Its 



if Jllltli 



a 
c 



>1 



1) 



1) 
>1 






o 
B 






o 
c 










o 
B 



o 
c 









a 

c 



o 
B 






a 
B 



a 

fl 







c 
B 






>> 






o 
c 



o 

c 






o 
c 



t 



Iftfif 



~ z * 



'Is 1 



o 
c 



1> 



o 
c 



8 






o 
c 



o 



o 
c 



a 
c 




a 
B 






o 
c 



o 
c 






g 



>1 






g 



o 
c 




o 
fl 



o 
c 



o 
c 



>1 



o 
c 






o 
c 



o 
c 



a 
B 



o 



ifiifilif 

IS SB ■ 



a 
B 



4> 



O 

c 






o 
c 



8 8 



o 
c 



o 
c 






If lalfllf 

aliliiiif 

~ z ? •="* 



o 

c 



o 
c 









o 

c 



>1 



SB 
>> 



** 



3 
C 

C 

o 



H 



ill nil! 



o 
B 



X ■ 












a 















8 







00 

9-" a S'l 

to )— to i_ 

VI Oi OO Oh 



a- 

U-. 

U 



a ai 



is* i 



V3 



=3 O 

33© 



S» •- 

1/3 CJ 



O 4) 

■c 'E 



o 

8 

i 



317 
comparisons of hydroperiod depth group frequencies. The significance of these 
hydrologic changes to swamp vegetation communities depends on species' tolerances 
(see Chapters 6 and 7). 

Trends in water levels and hydroperiods demonstrated with the hydrology model 
manipulations reflect the basin delineations determined with the topographic surface, 
vegetation maps, and water level recorder data. The Northwest basin represented by 
check stations at Billy's Lake, SCFSP, Suwannee River, Cravens Hammock, Sapling 
Prairie, Floyd's Prairie, and Transect 60, showed the greatest response to the sill's 
presence. Removal of the sill resulted in a downward shift of hydroperiod group 
frequencies of 1-3 classes (Figure 3-18). During low and average water levels, absence 
of the sill permitted greater fluctuations in water levels, and greater difference between 
the sill area and the remainder of the basin (Table 3-8). During high water conditions 
without the sill, this difference was not as great; the natural topography southwest of the 
sill restricts water flow from the area, creating a sill-like impoundment in the river 
floodplain as water leaves the swamp (Figure 2-20). The result is a condition similar to 
that of the natural sill at the southwest end of Billy's Lake and the constructed sill; the 
natural impoundment slows drainage from the area and creates pooling above the 
topographic rise or berm until water surface elevations exceed the crest, when overflow 
occurs. There is a similar berm southwest of Cypress Creek that impounds water in the 
river above it and may be creating the backwater effects noted in the Cypress Creek basin 
(Figure 3-15). 



318 



i 

i 

I 



3 

1 

I 

O 

"2 C/3 

B C 

CO Q 

<U •£ 

Si 

a ■ 

« "9 

II 
II 

s § 

m o 
00 o 

II 

p 
1= 

<£ i 

rfi o 

H. c 

u. c/1 

<U X 

© a 

— ° s 

o 2 

C o 

eo oo 

a. on 
g ~ 

00 ^ 

m a. 
o E 

x> g 





8 
ii 














































m 


T— 1 


00 


— 


— 


— 


— 


m 


•<*• 


«/■> 


m 


— 


•a) 


r- 


— 


NO 


«TN 








r^ 


o 


r» 


© 


© 


© 


© 


© 


NO 


© 


<* 


© 


© 


■* 


© 


NO 


ON 


a> 


I 

c 






n© 


o 


fN 


© 


© 


© 


© 


r- 


oo 


m 


On 


© 


© 


ON 


© 


<r> 


fN 








«n 


o 


r>- 


o 


© 


© 


© 


<r> 


© 


© 


■* 


© 


© 


o 


o 


o 


— 




» 






o 


d 


d 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


©' 


© 










fN 


CO 


«n 


00 


fN 


</"> 


On 


m 


© 


V) 


O 


NO 


oo 


fN 


fN 


fN 






1 

i 






r- 


T 


00 


— 


K 


fN 


3 


© 


T 


•^r 


© 


© 


fN 


fN 


00 


NO 


NO 








<N 


no 


O 


t-» 


00 


Tj" 


oo 


NO 


© 


<n 


^^ 


rf 


lO 


NO 


rj- 


^ 








«■- < 


00 


fN 


O 


NO 


On 


r- 


-3- 


*r\ 


© 


On 


U-l 


r-~ 


fN 


<n 


oo 


ro 








© 


d 


d 




© 


© 


© 


fN 




<ri 


© 


© 


© 


© 


© 


© 
i 


© 


■ JS 




- 

- 






































•dj 

II Gat 
Dept 




8! 


, , 


■* 


t*» 


r»i 


mm 


*Tl 


00 


(*• 


fN 


NO 


■* 


m 


^_ 


m 


m 


CO 


NO 


CO 


<k 


£ 


oo 


o 


NO 


*f 


t- 


On 


r*l 


«o 


CO 


«o 


On 


^^ 


ON 


oo 


rr 


r- 


© 


— 

Q 

1! 
> 


c 


a. 


fN 


00 


m 


NO 


fN 


^H 


T}- 


r-> 


v<^ 


ON 


fN 


00 


Tf 


r>- 




© 


F 


lint 


a 


O 


VO 


o 


ID 


NO 


■* 


, ^- 


■— < 


m 


On 


© 


>— ' 


«D 


o 


NO 




© 


11 1 


■ 

ft 


© 


d 


d 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


o 


©' 


© 


t5 


C 








































E 


1 








































■ 


a 






■*r 


no 


t 


NO 


■* 


iri 


Tf 


NO 


© 


fN 


<o 


m 


>0 


Tf 


1—^ 


^■H 


On 


■o 
bE 


> 






o 


*■< 


*— > 


fN 


© 


© 


•-< 


rn 


fN 


"• 


fN 


>— i 


— " 


© 


1—1 


— 


© 






d 


d 


© 


© 


© 


©' 


© 


© 


© 


©' 


d 


© 


© 


© 


© 


© 


© 


m - « 


B 

G 


c 
1 




00 


fN 


© 


t-- 






ON 




fN 


r- 


ON 


<n 


fN 


fN 


00 


OO 


m 


Aver 

Wat 

DeptI 


— 
so 


? 


m 


NO 


On 


« 


fN 


m 


"3- 


© 


fN 


On 


00 


© 


m 


© 


— 


m 


■* 




d 


d 


© 




© 


© 


© 






© 


© 






d 


© 


© 


© 


■ 










































E 

■ 


j 


5 




fN 


NO 

fN 


NO 

fN 


fN 


NO 

© 


fN 

00 


en 


*ri 


00 


m 


On 

•— 


© 

fN 


On 


fN 


NO 

fN 


NO 

fN 


fN 


c/> 










































L. 










































■ 










































** 




■ 






































1 £ 

° -a 

s 2 


> 


c 
1 
;c 

■5 

s 






u 

OB 

2 




X 






00 




00 

X 


u 

oo 

i 




00 


-o 

O 

o 




1) 
00 

CO 




00 

S 


w 1 

■ 




o 

u 




O 


> 


-C 

op 




E? 


o 


E 

> 


op 




6 
> 


JB 




o 


u. 
> 


X9 

00 


<u 


o 








-J 


< X > 


Q 


►J 


< X 


> 


< X 


> 


b 


< 


2 


> 




























Z 


















_ 






_ 


















o 








: 










o 










































o 






CO 








a 










S 








co 










? 






.s 

— 








CO 

o 
2 










o 
Z 
















a 








u 


































■ 








_^ 


































'-S 








« 


































1 








J 










































c/5 


























u 


V 






8 








V> 


























CO 

CO 


'C 






u 








s 


























U 









319 



T3 

0) 

3 

a 
a 

o 
(J 

I 
I 

00 

I 

CO 

I 







o 






CN 


CN 




t^ 


00 


o 


<n 


un 




00 




00 






<o 


<n 


m f- 






ll 






o 


O 


o 


r« 


fN 


00 


fN 


r^ 


o 


On 


o 


o 


>n 


o 


m 


o 


NO S 




a. 


& 

e 






o 


o 


o 


o 


On 


m 


TT 


»— * 


o 


^« 


o 


o 


oo 


o 


NO 


o 


On — 










o 


o 


o 


m 


— 


o 


oo 


o 


o 


--" 


o 


o 


VI 


o 


-" 


o 


m o 






£• 






© 


© 


© 


© 


O 


o 


O 


o 


o 


O 


o 


o 


d 


d 


d 


d 


©' o 












m 


r^ 


r-~ 


NO 


o 


m 


00 


fN 




o 


•n 


00 


■«3- 


oo 


<n 


00 


— o 






i 

58 






o 




K 


*"■ • 


n- 


r-~ 


00 


fN 


■* 


oo 


m 


f*"> 


V"> 


ir> 


r^ 


fN 


— 00 










>/-> 


Vl 


o 


>/") 


>r> 


no 


r- 




vO 


On 


o 


NO 


NO 


fN 


oo 


On 


NO fN 












IT) 


NO 


© 


</■> 


en 




fN 


r- 


On 


On 


r~» 


,— J 


O 


NO 


m 


00 tJ- 










o 


© 


© 


1 


O 


ci 
■ 


o 


O 


o 


o 


o 


fS 

1 


o 


fN 


d 


d 


o d 


£ 


jf 




•- 
4 








































a. 


J! 


re 


^~s 


o 


t^ 


no 


00 


fN 


On 


en 


^ 


fN 


On 


On 


o 


00 


rf 


On 


■* 


'j ^r 


Q 


f 


£ 


£ 


r- 


«f 


CI 


fN 


»-^ 


fN 


NO 


On 


m 


fN 


NO 


1—1 


V) 


Tf 


i—i 


i-M 


no >n 




u 

II 

> 


o 


Q. 


*— - 


<r 


00 


fN 


•<t 


ON 


V> 


m 


fN 


NO 


u-> 


>r, 


m 


w* 


»— 


00 


o r^ 


t en 


B 


2 


— < 


— " 


*n 


-™ 


— - 1 


On 


o 


o 


NO 


O 


■* 


fn 


o 


m 


O 


o 


o <n 


ii 




re 


© 


© 


o 


© 


© 


© 


o 


o 


o 


o 


O 


O 


d 


d 


d 


d 


d © 






1 






Tf 


in 


o 


m 


as 


00 


r^ 


o 


"• 




o 


fN 


m 


r» 


v-> 


NO 


On V-i 




I 








o 


o 


'— 1 


*- > 


© 


fN 


»— « 


«— ' 


•— < 


>— < 


fN 


fN 


— 


— 


•^ 


— 


O O 




8 






o 


© 


© 


© 


o 


O 


o 


o 


o 


o 


o" 


o 


d 


o 


d 


d 


© d 




^ 


Q 








































el 


■C 


.* 
2 


c 
o 






v© 




r^ 


fN 


NO 


r- 


o 


ON 


<n 


fN 


Tf 


fN 


r^ 


rn 


"T 


O Tf 


B 


•■c 

re 


? 


o 


O 


fN 


m 


fN 


«— i 


o 


<— 1 


fN 


NO 


O 


fN 


fN 


S 


On 




O »- 






© 


O 


o 


© 




© 


o 


o 


O 


o 








o 


d 




"" 




s. 

1 

1 


4. 
N 


'5' 




NO 

o 


fN 

00 




v> 


00 


m 


On 


o 

fN 


On 


fN 


NO 
fN 


NO 
fN 


(N 


NO 

o 


fN 

00 


»n 


V) oo 




M 












































- 












































| 




* 






































■ 


re 

re 

E 


~Z 

> 

1 


s 
c 

c 








u 

oo 




2 


u 

00 




t 

3 


T3 

T3 




00 




■a 

X 






u 

00 


•a 

s 


53 


■ 
e 
■ 




o 

u 




b 


o 


s 

> 


op 


0> 


E 

E 

> 


J3 
.00 




O 
O 




p 

> 


J2 
00 


b 

o 


b 




E 

6 
> 


J3 >. 

.SP B 




U 








Q 


J 


< X > 


<X> 


b 


hJ 


< I > 


Q 


J 


<X> 






















Z 












































o 




























1 






— 1 


















00 








_ 












2 






00 

o 










o 
















bo 

o 


















Z 










2 














2 










e 
o 

« 

i/2 


























i- 


<u 
















M 


























t« 


















Q 


























yi 


*C 
















I 


























u 


* j 











































U 


2 

Oh 















320 



-o 

3 

n 

1 
o 
o 

I 
I 

CX5 

•s 





II 




■«r 


t 






>/-> 


— Os 


o 




so 


— (*i 


m 




o — — 






so 


fN 


O 


o 


(N 


© © 


* 


o 


O 


O rf 


m 


(N 


SO o o 


eu 


1 
1 




■— • 


^— 


O 


o 


</-> 


o o 


o 


o 


o 


o o 


o 


m 


— o o 






— 


O 


o 


o 


o 


o o 


<N 


o 


o 


O rf 


o 


o 


— o o 






© 


© 


© 


o 


o 


© © 


O 


o 


o 


O © 


o 


o" 


o o o 








-t 


oo 


o 


so 


o 


f* £ 


fS 


00 


00 


SO — 


o 


SO 


00 00 o 




e 




o 


os 


m 


<N 


m 


CN Os 


K 


SO 


»— * 


m — 


SO 


</"> 


— so O 






(N 


tn 


o 


o 


Os 


*** ft 


<^ 




SO 


m so 


TT 


rsi 


oo r-> Tj- 




1 




<N 


«— i 


F 


Os 


<r> 


oo P"; 


rn 


V) 


Os 


en m 


*n 


SO 


oo >n r<- 






i 




o 


© 


© 


©<? 


O 




o 


o o 


o 


i 


o o o 


IV .e 
































.dj 
II Gal 

Dept 


■* OS 
IS 


, , 


00 


o 


On 


r- 


o 


so m 


(N 


o 


<N 


SO 00 


o 


Os 


m wi m 


£ 


m 


«n 


^— 


sO 


f- 


CN 00 


<n 


en 


SO 


r> «n 




rr 


00 00 — 


■ J 


n. 


en 


r» 


*— i 


ir> 


1-* 


00 Tj- 


m 


o 


CN 


Os •— • 


V-) 


Os 


oo in rf 


XieS 


0/ 


os 


rM 


CN 


TT 


— J 


so rn 


o 


v-i 


" — 


o o 


r>- 


Os 


ONt 


11 1 


© 


o 


© 


© 


© 


o o 


o" 


o 


O 


o o 


o 


o 


O o o 


1 






o 


ON 


«/-> 


O 


SO 


SO CN 


m 


o 


Os 


m TJ- 


VI 


Tf 


— — so 


^ 






»— ' 


o 


>— ' 


rsi 


o 


r-H fNJ 


»— i 


•— < 


O 


— o 


o 






1 


1 




© 


© 


o 


© 


© 


© © 


o 


o 


d 


o o" 


o 


o 


d d d 


eS 


° 




































CN 


00 


r« 


ro 


« 


so rn 


w-> 


m 


oo 


rv w« 


oo 


r-~ 


— ,— so 


1 * m. 


fc C 


? 


s© 


so 


OS 


CN 


fO 


SO 00 


00 


rf 


>o 


t-» r> 


00 


m 


tj- so oo 


I*fe« 




© 


© 


© 




o 


o o 


o 


o 


o 


o o 


o 


o 


odd 


i 

E 


Ms 




m 


Os 


o 
rsi 


Os 


CN 


gs 


(N 


SO 

o 


oo 




00 


m 


as ° T *' 

g fN Os 


09 
































u 
































V 
































** 


■ 






























SS U 


c 
_ e 




u 

00 




■a 

X 






00 


00 

a 






u 

00 

CO 


00 

X 


00 

CO 


igh 

ery High 
ooded 


s 1 






6 

> 


J2 
00 




O 

_o 


o 


P "a 


<u 


b 


o 


> .SP 




B 

> 


o 






< X > 


fcu 


J 


<x> 


Q 


-J 


< 3G > 


< 


X > u. 








£ 






















^ 










o 






















o 






J 




c 








S3 














B3 






j 




1 








£ 






j3 








3 






2 




O 
o 








5 






O 








o 

o 

z 




e 
I 














£ 


















« 














on 
































— 


















8 














u 


















M 














,u 


















B 














— 


















■ 














«E« 


















J 














O 

































U 



















321 



-a 

<L> 
3 
B 

c 
o 
o 

I 
I 

oo 

I 
CO 

u 
X> 

a 
H 



















































o 






00 


(M 


o 


_ 


_ 


_ 


M 


o 


o 


o 


Tf 


— 


pa( 


>ri 


1-H 


fN 


00 






II 






r> 


o 


o 


o 


o 


O 


o 


■* 


*n 


00 


00 


o 


o 


NO 


o 


m 


m 




a. 


i. 






o 


o 


00 


o 


o 


o 


o 


(N 


*— 


en 


On 


o 


o 


On 


o 


© 


NO 






— 

o 






no 


o 


o 


o 


o 


o 


o 


t*> 


o 


o 


, ^- 


o 


o 


•-" 


o 


© 


(N 












© 


© 


© 


© 


o" 


o 


o 


O 


o 


d 


d 


d 


d 


o 


d 


© 


© 












rj- 


CM 


o 


*f 


On 


00 


m 


o 


o 


r-~ 


NO 


ON 


00 


O 


</■) 


!•«• 


>^> 






1 

o 
55 






m 




o 


r- 


On 


CN 




*"• 


oo 


o 


t 




ON 


o 


* 


NO 


NO 










on 


CN 


m 


00 


ro 


t^ 


r^ 


o 


■<* 


nO 


o 


(N 


tn 


Cl 


(N 


(N 


(N 










*^ 


On 


on 


ON 


r- 


© 


On 


00 


CN] 


fN 


OS 


U-l 


00 


o 


■* 


On 


V^ 










© 


© 


o 




© 




O 


© 

1 




i 


d 


d 


O 


1 


© 


rn 
i 


© 
i 


£ 


j= 




- 

4. 






































I 

JO 


B 
Q. 


E 


1 





m 


VI 


_ 


r> 


r> 


o 


t"» 


t^ 


00 


On 


oo 


r- 


__ 


_ 


^_ 


m 


*— 


I 


•5 


£ 


»— 


f» 


^— 


rr 


>n 




5- 


On 


NO 


<N 


ON 


<n 


■<* 


i— H 


NO 


On 


NO 


■c «« 


-c 




cL 


m 


•>*• 


00 


CN 


* 


Tt 


r- 


NO 


On 


ON 


fN 


— 


o 


m 


r- 


F» 


^^ 


iTc« 


L. 

H 


U 
ii 
> 


c 
o 


i. 


o 


r^ 


o 


NO 


m 


■* 


K 


rv) 


«o 


ON 


O 


^~ 


TT 


o 


© 


rs 


© 


ii 


■S 
■ 

* 


1 
■ 

2 


— 


© 


© 


© 


© 


© 


o 


o 


o 


d 


d 


d 


d 


d 


d 


© 


© 


© 




■c 


e 










































E 


1 










































i 


a 






r> 


NO 


CO 


m 


NO 


<T> 


m 


00 


m 


NO 


<n 


v^ 


On 


, ^- 


m 


<o 


1— H 




| 


■ 






o 


•— > 


»-< 


CN 


o 


© 


T— < 


*— < 


«— i 


m 


(N 


•— ' 


i— ' 


^~ 


<N 


m 


CM 










© 


© 


© 


© 


© 


© 


o 


o 


o" 


d 


d 


O 


d 


d 


© 


© 


© 




8 


2 












































^ 
^ 


c 
1 




CN 


t*. 


CN 


-<t 


CN 


cn 


</-> 




m 


On 




NO 


«r> 


t^ 


ro 


00 


m 


8 « 


i- 


a 


? 


CN 


^r 


00 


"i 


>— I 


CN 


-* 


p 


«N 


NO 


NO 


t> 


o 


«ri 


t^ 


NO 


>TN 


3* 




I 

*- 




O 


© 


© 




© 


© 


d 






d 


d 


o 




d 


© 


© 


© 




■ 












































E 


| 


3 




CN 


no 
cn 


NO 
CN 


CN 


NO 

o 


CN 

00 




<^ 


00 


m 


ON 


o 


On 


<N 


NO 


NO 


<N 




n 












































k 












































1 
















































■ 






































■ 


3 


1 

E 

I 

1 
B 


1 

> 


c 

'■5 

c 
o 

u 




o 


n 

bo 

E 

s 

> 


op 


-a 

X 

b 


b 


o 


u 

2 

6 

> 


X! 


-5) 
SB 
b 


u 

Ofi 

S 
c 

> 


op 


•a 

s 


T3 

<u 
T3 
O 

o 


o 


00 

s 

> 


bo 


■a 

X 

b 




u 








- 


< I 


> 


Q 


J 


< X > 


< 3C > 


Lu 


- 


< 


X 


> 






























15 

o 




















"I 






a 
























•a 












■o 
i 






£i 








>—« 










3 








eo 












2 






x: 








i/5 










O 
o 

2 


















s 
o 
•r 










.M 


































■ 








M 


o 


































«- 
&C 
« 








"c 


o 
£ 
























(A 


_<: 








8 








Crav 
Ham 
























Cypr 
Cree 













322 



§ 

C 
I 

o 



00 
I 

en 

2 

« 





it 




















</"l CN 


SO 00 — 


SO 00 OS «— i 


_ 


— rn — < 


i- — — Os O 






O SO 


CN 00 O 


en en oo o 


r- 


O so O 


© © © CN — 


A. 


1 




O tJ- 


O ON o 


SO o o o 


o 


O CN O 


O O O 00 o 






© o 


o o o 


o -* © o 


o 


o o o 


O O O </-i o 




Cfl, 




o o 


o o o 


o o o o 


o 


o o o 


o o o o o 








en cn 


o o o 


2> -3- r*> so 


o 


(N ^ (N 


so en so O O 




1, 




o so 


m r^ cn 


C OS SO CN 

2 i> t* os 


Tt 


vO © o 


OS tt — — i so 




e 




en — 


tt r* <n 


o 


OS ^ "S- 


OS CN so 00 rj- 




03 




00 OS 


rn m so 


^ 00 en tT 


en 


00 O CN 


«r> O v^ en r» 








© o 


o o o 


t o o o 


o 


O T mm 


o — o o o 


<v js 


















*dj 

II Gal 
Dept 




,_ 


O ON 


-oo 


mm OS SO SO 


00 


^ so oo 


r- t*» so r- — 


-e 


en so 


as -t -h 


o r^ oo <n 


en 


r*« «n r~ 


o <n os tj- os 


> | 
m* 


& 


o en 


v-> en u-> 


00 OS ^ i— 


TT 


cn >/■> en 


en cn un oo cn 


T. E £ 


a 


— i o 


O ViO\ 


Os O O «— 


CN 


00 — so 


«o V> Tf •— 00 


1 i 


o o 


o o o 


o o o o 


O 


o o o 


o o o o o 


I 


c 
e 
















■ 


« 




CN TT 


SO CM SO 


CN O SO mm 


CN 


>/-> CN </-) 


rf «o o so t-- 


"O 


1 




l-H I— 1 


*-* © o 


m<* <— i <— i CN 


O 


»— < <— ' •— i 


o o — o o 


I 




o o 


o o o 


o o o o 


o 


o o o 


o o o o o 


8 


a 
















m h " 


.* = 
















2*« 


B p 


.— s 


»— rn 


oo os m 


oo -«3- f~ rn 


mm 


CN en CN 


^f <o SO 00 SO 


4, X 

Off 


E 


— CN 


en CN Tf 


O mm en «r> 


© 


CN ^t SO 


o — en *n r-- 


•< ^ o 




o o 


o o o 


o o o o 


o 


o o o 


o o o o o 


■ 


















a 

E 

1 


Is 

en w 




Is 


1/1 

en wn oo 


en g W * 


CN 


3 *> ~ 

S CN CN 


^> fO "O 

o £; en >/n oo 


bB 


















h 


















■ 


■ 
















Sill Gate 
General Wa 


e 
_ o 

si 

u 




Q J 


Average 
High 
Very High 


Average 
High 

Very High 
Flooded 


O 


Average 
High 
Very High 


Dry 
Low 
Average 
High 
Very High 












£ 










T3 








o 


— 








e 




Z3 




■g 


on 




— « 




s 




00 

o 
2 




1 






on 

o 
2 


c 
e 

■ 

ft 












u 






Check 












1 

O 

Q 


B 

M 

2 





323 



4> 

3 

a 

'•♦3 
1 
O 

o 

I 

00 

I 

CO 

CO 

H 





_ 
























o 
II 




rfr OS <n <— 


— — r-. 


00 


_ 


_ 


— l> 


en 


SO O 


so — 






so cn r- o 


o o <n 


^^ 


o 


o 


O Tf 


1—1 


00 oo 


— © 


fix 


e 




CN VI so © 


o o —> 


en 


o 


o 


O en 


«n 


CN CN 


© o 






© r» © © 


o o o 


o 


o 


o 


O V) 


en 


O Tf 


© © 




B 




© © ©' © 


do © 


© 


© 


d 


o o 


o 


d d 


©' © 








S5 r^ r- os 
S <n S en 


Os en £ 


Os 


r-» 


Tt 


T CN 


o 


2 °° 


>n — 




| 

c 




~ w £ 


00 


o 


■<1- 


v-i so 


"4- 


£ ^ 


en © 






)§ O rt On 


Os so £ 


— 


r^ 


r-> 


00 fN 




?S "4 


00 00 




35 




^ en cn so 


ooo^ 


en 


CN 


o 


<n cn 


CN 


CS vo 


en r~ 








*? © o © 


-J o -7 


© 






o o 


d 


f? d 


© ©' 


4> x 


■ 






















adj 

II Gat 
Dept 


■a! « 


M 


so <n oo oo 


m "ri Tf 


Tf 


CN 


o 


en m 


Os 


O en 


<*» — . 


I £ 4 


SO CM Os l> 


en tj- r^ 


Os 


«n 


00 


TJ- SO 


CN 


so Os 


en so 


x ^ 

' -B. 
> I 

■ 


& 


Os <n — so 


rtOOO 


F» 


Os 


OS 


OS Tt 


o 


OS — i 


r- r^ 


t m t 


C 


os o o en 


so F — 


'— , 


>♦ 


CN 


en — 


o 


Os O 


© ■* 




© © © ©' 


© © © 


© 


© 


d 


o d 


d 


d © 


©' © 


"O 


e 






















E 


e 






















«s 


« 




r*» oo <n f» 


in so so 


r- 


Os 


r>. 


en en 


•n 


Os <n 


tt r^ 


T3 
I 


I 




CN «-* ~h •— i 


O «-i •-] 


o 


o 


o 


«n o 


O 


1— < <— 1 








© © o © 


© o d 


© 


© 


o 


d d 


d 


© ©' 


© © 


as 


a 






















« 5 X 






so os >n r^ 


»— 1 t**« ^H 


i> 


00 




CN O 


r» 


© r-- 


— i en 




? 


— o — < en 


en «n oo 


00 


en 


<o 


S oo 


00 


rr en 


>n r^ 


2r !^ »■■ 


^35 




© o o O 


© o © 


© 


d 


o 


o d 


o 


©' © 


© ©' 


41 

E 

■ 






os S £ 
en {J cn os 


"1 3 * 

CN « CN 


CN 


so 

o 


CN 

00 


»n 
en <n 


oo 


« s 


© Tf 
(N Os 


as 
























k 
























1 


■ 






















Sill Gate 
lencral Wa 


c 
_ e 

u 




verage 
igh 

ery High 
looded 


H 


2 


b 




u 

00 
cS 

r x: 


■a 

X 

ft 


00 

B 


ery High 
ooded 


V? 






< X > E 


J < s > 


Q 


-1 


< X > 


< I > E 








£ 














^ 






T5 

e 




o 

3 


1— -: 




,^ 








o 
1 






S 




o 

O 

2 


1 




S5 
o 

2 








O 
o 
2 




a 














e 
























OS 
























2 








C <D 
















■M 
I 








1 "C 

B "3 
















■1 








3 g 

























Q £ 

















324 



T3 

0> 

3 

s 

E 
o 
u 

I 

oo 

I 

en 
u 

1 













































o 






n© 


PM 


ON 


— 


pa 


_ 


_4 


"5J- 


CM 


ON 


'* 


,-H 


pa 


, — i 


pa 


ON pm 




II 








o 


o 


© 


© 


o 


o 


co 


■* 


p— 1 


CM 


o 


O 


NO 


O 


00 O 


Gm 


1 

o 






t^ 


o 


o 


o 


O 


© 


o 


On 


O 


m 


CM 


o 


O 


NO 


O 


en o 








o 


o 


CN 


o 


© 


o 


© 


en 


o 


o 


^~ 


o 


O 


—" 


O 


o o 




55 






© 


© 


o 


© 


© 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d d 










00 


T—H 


o 


00 


r- 


NO 


r-^ 


CO 


o 


ON 


r^- 


o 


CM 


u-> 


00 


£ £ 










o 


«o 


cm 


"fr 


On 


On 


00 


en 


oo 


o 


On 


en 


«n 


>n 


a— « 


en ^ 








no 


pa 


T 


00 


NO 


en 


CM 


en 


c- 


*~* 


00 


o 


On 


On 


t^ 


X * 




n 






en 


oo 


en 


■<3- 


oo 


— 


r« 


en 


v> 


r^ 


<— i 


r^ 


f-- 


en 


ON 


On vc> 








o 


O 


o 


© 


©' 




d 


d 


d 


CM 




d 


d 


d 


d 


9 o 


4- xT 




«y 




































•dj 

II Gal 
Dept 


i 


a- 
R 


^_^ 


NO 


oo 


■* 


o 


(N 


t-» 


NO 


ON 


ON 


f- 


C*» 


*-H 


o 


r^ 


ON 


CM en 


£ 


£ 


NO 


CM 


00 


NO 


r>- 


-— 


CM 


CM 




NO 


en 


NO 


CM 


t-a 


r- 


— o 


— 




EL 


ON 


O 


CM 


>r> 


cs 


o 


ON 


O 


en 


00 


00 


•* 


«/-> 


, ^- 


r- 


en 'sr 


iKi 


u 
> 


C 
O 


«y 

a 


o 


r^ 


o 


</■> 


NO 


C<1 


NO 


o 


r«- 


ON 


o 


en 


NO 


o 


r-- 


— <r> 


I 'S 


09 


© 


© 


© 


O 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d d 


tj 


& 






































E 


o 






































I 


s 






■* 


«n 


f- 


NO 


</■> 


r- 


<-H 


en 


NO 


en 


«n 


CM 


«r^ 


<n 


r~ 


PM ON 


I 


cs 






o 


<— i 


o 


o 


o 


o 


•—1 


o 


o 


CM 


»—i 


•— > 


— - 


o 


<— i 


•X O 








o 


© 


© 


© 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d d 


55 


o 










































e 
o 




On 


On 


NO 


O 


NO 


On 




en 


o 


TT 


•n 


r- 


r-- 


NO 


On 


•M T5- 


Aver; 

Wat 
DeptI 

Che 


BO 


? 


O 


en 


NO 


00 


o 


*■" 


■* 


NO 


00 


CM 


CM 


■el- 


r- 


CM 


<T) 


On O 




© 


© 


© 


O 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d *-' 


_u 








































"EL 
E 


H 

x 


~s 




CN 


•<* 

SO 
CM 


NO 


CM 


NO 

o 


CN 

00 


en 


tr> 


oo 


en 


On 


o 

CM 


S 1 

On 


CM 


NO 
CM 


NO *- 

CM CM 


BO 








































Lj 








































5 




■ 




































= £ 


1 


e 
o 
a 

s 






a 

oo 

c3 




■a 

X 






u 

60 

a 




■a 

X 


u 

oo 




-a 

5 


T3 

OJ 
T3 
O 
O 




D 

00 

C3 


5 


8 1 

9 




g 




o 


> 


00 




fr 


o 


E 
> 


.£P 




E 
> 




0> 


o 


E 
B 
> 


•S r*N 

.2P u 


u 








J 


< X > 


Q 


-1 


< ffi > 


< I > 


E 


< X > 




























o 

IS 
















- 






— 
























S3 








o 






i/3 








a 










3 








S 








2 






5* 








on 










O 
o 
Z 








^ 
£ 






Check Station 


<*> 


c 

'3 
























| 

CO 


•a 




U. 


Oh 
























o 


J 













































325 



<u 
3 
C 

i 

o 

o 
■ 

00 
I 

m 

u 

I 

H 



p 

(Slope = 0) 


- t - rf fS 

o (N o -3- o 

o o o m o 
o o o — o 

© O © © © 


t 00 - - 

NO o o o 

<— < oo o o 

— TT O O 

o o o o 


0«/)Oi 
On O M O 
<n O O f* 

o o o o 
© o o © 


— t"» <0 NO — « 
O *3" O rf o 

©moo© 

© © © <N © 

© ©'©'©'©' 


| 

o 

53 


irin h oo o 
t n * m \o 
o\ — — r- — < 
-3- on r- t> r* 

»■« o © © © 


-1.9526 
0.6120 
0.5488 
0.9867 


0.7482 
0.8027 
-2.7866 
0.4655 


rr o r- rr o 

O On M ^t t 

on oo m ^ r^ 

—' © ©' © © 


(X=Sill Gate 
Water Depth, 

Y=Check 
Station Water 


Q. 

Q 


<N o vo v-i r» 

oo oo t-~ © <n 
r- on m rr o 

^ O t t ON 

© o © o © 


00 Tf O 00 

m r^ ^r o 
m rs no ■* 

On O — SO 

© o © © 


n vitet 

On no m f^ 
O O On — 
— U1 (N — 

o © o o 


no r- r^ o m 

m (N o r«- — 

O Tf M 00 >-< 

m o © r-j On 
© O O © © 


■2 1 
If 


O On fl VI VO 

— O *-* © O 

© o o ©' © 


p»- vo m oo 
© o o o 


r- r- ^r cn| 

O <-< CN «-h 

© © © o 


•^- «n m m no 

— ** — o o 

o" © o o o 


Average 
Water 

Depth at 
Check 
Station 


f 


00 m f^ On On 

m </-> r-- oo o 

O O © © — i 


— m On O 
m rn tj- oo 

©'©'©'©' 


NO rn 00 On 

o t mv> 

© © © © 


o t N wn n 
CM rn <n Tf vo 
o ©'©'©'© 


Sample 
Size 

(n) 


iS rs ^ 

© ojj m <n oo 


on ° 2f 

f» S N On 


*r\ S *° *"' 

<N ^ CM <N 


£ rM ^ 

o g m iri oo 


Sill Gate 

General Water 

Level 

Condition' 


Dry 
Low 
Average 
High 
Very High 


Average 
High 

Very High 
Flooded 


Low 
Average 
High 
Very High 


Dry 
Low 
Average 
High 
Very High 


J 


j/5 
o 
2 


o 

I 
1 


bo 


1 


B 
1 
■G 
■ 

en 
ji 

B 

B 








Moonshine 
Ridge 





326 



•a 
&> 

§ 
1 

o 

O 
i 
i 

00 

I 

m 

1 





_ 








































o 






o 


m 


^_ 


pa 


o 


,— i 


m 


— 


_ 


— 


— . 


o 


r- 


NO 


— 


— 1—1 




II 






00 


00 


o 


O 


NO 


O 


■* 


© 


o 


o 


o 


m 


(N 


00 


O 


o o 


CM 


1 






m 


m 


o 


O 


O 


o 


o 


o 


o 


o 


o 


m 


o 


^^ 


o 


o o 








o 


<N 


o 


o 


O 


o 


o 


o 


o 


o 


o 


r- 


o 


o 


o 


o o 




s. 






o 


© 


© 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d d 










2j 


TT 






i> 


00 


o 


o 


o 


o 


c-> 


Ti- 


o 




NO 


O NO 




1 






o 


00 


o 


o 


(N 


1— < 


On 


00 


On 


m 


m 


— 


o 


(N 


NO 


m on 








<^i 


TT 


00 


o 


O 


o 


On 


o 


<n 


o 


m 


rj 


<N 


NO 


ON 


00 o 




Sa 






"■"I 


l> 


t^ 


r»- 


rr 


00 


00 


m 


ON 


•- 1 


00 


m 


O 


ON 


>— 1 


00 On 








i 


O 


© 


d 


d 


d 


d 




d 




d 


d 








d d 


K JS 




B 




































.dj 

II Gal 
Dept 


J! 
1 


8! 


^^ 


On 


o 


•* 


00 


m 


00 


m 


tj- 


00 


</-) 


,_ 


in 


CS 


m 


in 


O NO 


£ * 


(N 


o 


m 


f> 


m 


OS 


^H 


# 


in 


1-H 


"* 


CM 


r^ 


00 


v> 


(N 00 


□ 

II 

> 


C 
■ 


B. 


on 


rs 


© 


m 


m 


ON 


NO 


NO 


On 


r- 




00 


NO 


On 


m 


O NO 


ii 1 


«y 


On 


o 


TT 


m 


CS 


00 


(N 


o 


oo 


r» 


On 


<N 


t^ 


On 


oo 


ON ON 


■s 
n 


O 


o 


© 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d d 


■o 


s 






































E 


o 






































■ 


•c 






00 


m 


<N 


00 


c-i 


ro 


00 


NO 


m 


TT 


i— * 


00 


o 


NO 


NO 


ON "*f 




1 






•— • 


•— i 


•" > 


'-* 


O 


>— • 


o 


•— 1 


o 


O 


>— i 


o 


i-" 


»■* 


o 


o — < 








© 


o 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d 


d d 


n 


O 






































Average 

Water 

Depth at 




c 
e 




© 


ON 


t- 




on 


m 


NO 


ON 


00 




rn 


■* 


*o 


fN 


*— M 


in r~- 


c 




? 


t^— 4 


o 


m 


NO 


* 


l> 


© 


cs 


cs 


■* 


NO 


o 


<N 


O 


CS 


in oo 




© 


o 


d 


d 


d 


d 






d 


d 


d 












M 








































a 

E 

■ 


1 






m 


On 


o 

CN 


on 


«n 
cs 


n© 


NO 
CS 


CS 


NO 

o 


CS 

00 


«n 


«n 


00 


m 


On 


O rr 
CM ON 


oB 








































h 








































V 








































Sill Gate 
General Wa 


1 

.2 


s 
o 
•c 

? 

e 
o 

u 




n 

00 

2 
G 
> 


x 
ep 


■a 

53 


73 

-v 
o 
o 




U 

00 

2 

> 


X! 

op 


■a 

X 


fr 




u 

00 

s 

E 

6 
> 


x: 
op 


•a 

a 

4> 


u 

00 

CO 
E 
U 
> 


XI 
.60 


ery High 
ooded 








< ffi > 


u. 


-J 


< EC > 


Q 


J 


< X > 


< DC > E 










£ 


























^ 














o 


























o 








"5 

o 






a 

3 








^1 
00 








— 










1 








5 






o 

o 








1 








GO 

O 

2 










o 
1 






■ 
e 
■c 

08 

■3 
















8 
























eS 
















s 
























■k 
















5 


h. 






















B 
















3 


> 






















u 
















09 


5 























327 



"O 

<L> 

3 
_C 

C 

o 



00 
I 

X> 

H 





© 
II 
























m — "O *— 


— 


_ 


«— in on 


O 00 On — 


ON 


M 


NO — 






NO © CN © 


o 


o 


o m m 


00 (N Tf O 


■* 


o 


r~ o 


c 


1. 




r> o on o 


o 


o 


o o o 


m *- no o 


<^> 


o 


m o 






ro © rr © 


o 


o 


O On o 


o 00 o o 


^~ 


o 


o o 




K 




© O © © 


© 


© 


o © © 


© © o © 


o" 


o 


d o 








© no — Tf 


m 


OS 


ON CO © 


J^ NO 00 -N 


ON 


o 


o t^- 




4* 

& 

e 




m oo no — 


CN 


On 


— I-- © 


2 t^ — r- 


p» 


F 


*n -h 






On rf vo <3\ 


K 


•«*■ 


00 00 o 


3C «n -^- «-> 


r~- 


rf 


— o 




K 




— oo — r*» 


<N 


■"I 


no o r^- 


®N«h 


<N 


NO 


^f CO 








odd © 


O 




o o o 


•? O O O 


■ 


O 




*> JS 






















.dj 

II Gal 
Dept 


-* B 

if 


*~* 


oo o o m 


NO 


o 


On >0 On 


ON fN T Vt 


<N 


00 


m rn 


£ 


r- no — IS 


O 


1—1 


<r> <n r^ 


<N >T) O »— 


<r> 


fN 


m o 


> % 


O. 


O M (N * 


o 


^^ 


*T) (N fj 


On m (N <N 


*r> 


r^ 


m ^o 


T- K S 


4. 

O 


o oo o r-- 


CN 


m 


no co r>- 


On O O tT 


o 


«N 


— <n 




o © © o 


© 


© 


© © © 


O O O O 


o 


o 


d o" 


■2 


B 

e 




















eg 


X 




t ^- * a 


co 


t-» 


0\0h 


— o "* r-- 


m 


On 


r- r- 


c 
I 


B 




o — o o 


© 


o 


-« o o 


n (N «- 


f^M 






'5 




© © © © 


o 


o 


© © © 


o o o o 


o 


O 


o d 


8 


a 






















u 9 




m oo c> o 




1— 1 


cn o t*- 


oo oo —i r^ 


»-N 


00 


00 NO 


Eif 


■ 


o cn <o r^ 


o 


<— 1 


d NO IS 


— > o — m 


CO 


v-> 


00 — < 


< ^ © 


U$5 




© © © © 


© 


© 


o d o 


o o o o 


o 


o 


O — ' 


a 






















a. 

E 

■ 


3 c 

1/3 w 




^ s « - 

M J^ fN N 


NO 

o 


CN 

00 


co </-> oo 


ft S n <^ 




1 

NO 
(N 


NO — 
fN OJ 


BQ 






















u 






















■ 






















** 


a 




















II Gate 
ralWa 


e 
_ c 

SI. 




o -Sir 








> .S 1 o> o 
< I > E 




u 

00 
CO 


-a 


3 


© 




2 < x > 


Q 


o 


ft "S r*N 

> .2? B 
< X > 


o 


It I 1 

< X > 
















^ 










■5 




a 








o 
a 


— 








e 




CO 


a 






3 


CO 








S 






CO 

o 
2 






o 

o 

Z 


• ••* 






■ 
o 
















.«> 






m 
















. 






S 
















~* 






V) 






M « 










~ 






M 
| 








CO Oh 










| 

co 







328 



C 

I 

o 
o 

I 
I 

oo 

I 

CO 

JD 

3 

as 
H 







■g^ 












































© 
ii 






— 


oo 


M 


co 


oo 


o 


CO 


r- 


f^ 


Ov 


^> 


SO 


^* 


— 


>— 


_ 


V0 00 










o 


00 


o 


co 


vo 


00 


co 


CN 


o 


00 


o 


00 


o 


o 


o 


o 


— Tf 




a. 


a- 

a 
I 






o 


SO 


o 


ON 


CN 


CO 


vo 




o 


1— H 


o 


VO 


o 


o 


o 


o 


— so 










o 


© 


o 


vo 


o 


o 


r-- 


O 


o 


Tf 


o 


r^ 


o 


o 


o 


o 


SO O 






E 






© 


© 


© 


© 


© 


© 


© 


O 


o 


d 


d 


d 


d 


d 


d 


d 


d d 












© 


-* 


r- 


vo 


o 


o 


CN 


VO 


vO 


i— < 


VO 




00 


VO 




r-> 


2? <=> 






| 

o 
■ 






co 


vo 


r^ 


cs 


00 


VO 


co 


•* 


co 


o 


vO 


co 


o 


VO 


SO 


r- 


)°. °° 










CN 


00 


co 


00 


(N 


rf 


CN 


vo 


o 


o 


CN 


V) 


VO 


CN 


VO 


00 


S n 










© 


VO 


rn 


•"* '. 




co 


r~ 


•* 


00 


CN 


Os 


— 


00 


r^ 


o 


t^ 


• "*. 












© 




i 




©5 


© 


© 


o 


d 


d 


d 




d 


r— * 


d 


i 


i 


£ 










































1 


"EL 


J* 




> _ v 


00 


oo 


cn 


co 


VO 


Ov 


VI 


V0 


CN 


VO 


VO 


oo 


ON 


T 


ON 


00 


O ON 


i 


4- 


£ 


x 


— 


00 


o 


CN 


CN 


CN 


CN 


CO 


»— < 


co 


o 


r- 


ON 


vo 


VO 


o 


vo oo 


13 — 


£ 




c. 


CN 


(N 


"3- 


ON 


CN 


On 


VO 


■<* 


^^ 


»— 1 


oo 


co 


o 


-!t 


t^ 


CN 


O SO 


T."cJ3 


h 

H 


II 

> 


C 
o 


«, 


CN 


o 


vo 




V0 


Ov 


© 


o 


CN 


o 


r» 


o 


VO 


SO 


Tf 


VO 


CN CO 


II 


1 


•a 


C 


O 


© 


© 


© 


o 


© 


© 


o 


O 


d 


d 


d 


d 


d 


d 


d 


d d 




I 


e 
o 










































■ 


9 






o 


«— t 


cn 


00 


CM 


ON 


co 


o 


VO 


t 


VO 


CO 


rr 


rr 


V0 


co 


— 00 




■D 


■ 






*-H 


^^ 


CN 


P"l 


«-^ 


r- 


TT 


CN 


CN 


o 


<** 


"—I 


CN 


o 


o 


"- — 


co — i 




I 


4* 






© 


© 


© 


© 


© 


© 


© 


o 


O 


d 


d 


d 


d 


d 


d 


d 


d d 




■ 


a 








































£ 1 

4* 


■ 




c 
e 




o 




vO 


r- 


co 


vo 


00 


oo 


vO 


ON 


VO 


vO 


CN 


ON 


o 


o 


00 ~ 


f 


a 


? 




CN 


t 


— 


CO 


Tt 


f— t 


CN 


<r> 


CO 


vO 


ON 


CN 


V— H 


co 


VO 


On CN 




in 


© 


o 


o" 


"* ' 


^ H 


© 


© 


O 


O 


d 


d 


d 


^^ 


d 


d 


d 


©' ^ 




a. 
E 


J 

9 


c" 




so 

o 


cn 

00 


vo 

co 


V0 


oo 


CO 


ON 
*— < 


o 

CN 


Os 


VO 
CN 


VO 
CN 


SO 
CN 


CN 


so 

o 


CN 

00 


VO 
CO 


VO 00 




aS 












































L 












































B 












































*■» 




■ 






































■ 


■ 
* 


"1 


c 
o 

■c 








n 




-§> 


o 




X! 
00 


T3 
(U 

-a 
o 

O 




U 




00 






u 


•§) 


1 
E 


s 


■5 

c 








00 




X 


1 




2 




00 




s 






60 

9 


£ 


X 


■ 

e 
■ 


© 

a 




b 


o 


E 
G 

> 


x: 
■ S? 




E 

<u 
> 


s& 


£ 

3 


E 
U 

> 


x; 
op 




£> 


o 


P 

> 


x! ^ 













Q 


J 


< X > 


< I > 


E 


< X> 


a 


-; 


< X > 






















£ 












































o 








__ 




















c 












































V 






— 










^3 








00 








— 












o 










































5 






00 

o 










o 








£ 
^ 








00 

o 


















z 










Z 














Z 










c 
■ 
■s 












































I 












































** 












































H 


























- 


















M 


























00 


















t 


























u_ 


















X 


























u 


















U 


























oo 

















329 



3 



C 
O 

I 
I 

oo 

I 

.O 
CO 

H 







>— V 




























e 






o m — — 


o 


*-* m r- 


^H 


^ 


so rn 


Os 


SO 


"4- 00 — < 






II 






oo fN o © 


o 


o — t^ 


o 


o 


vo r~- 


r~ 


>n 


«n in o 






ft. | 






m fN © © 


^r 


o o >o 


o 


o 


V) 'T 


so 


fN 


Os O O 










o r*i o o 


^N 


o o o 


o 


o 


On ** 


*■- 


•*" ' 


sO O O 






K 






© © © © 


© 


© © © 


o 


o 


O O 


o 


O 


O O O 












2^ ^ r<^ Tf 


r<"> 


<* £ — 


o 


O 


t^ VO 


o 


00 


f»- — o 






e 






££ O t"» 00 


O 


o £ £ 


■<f 


(N 


Tt" <^l 


SO 


m 


m so os 










£J oo m oo 


so 


fv. •— V) 


r- 


<^l 


O <^ 


00 


o 


O rn o 










NO\\Oh 


fN 


t "^ * 


t^ 


r^ 


o o 


—* 


l> 


fN fN ■<* 










<?' o ©' © 


© 


©<? 9 


o 


o 


o •-< 


o 




o o o 




K 


-e 


- 
























C5 


Ii4) 




_ 


OMNOM 


t^ 


o ^ «o 


o 


oo 


o o 


W-l 


|M 


O OS o 




w 


A 1 


* 


£ 


fN fN Os V> 


fN 


t*> o m 


r- 


t-» 


O r<1 


(N 


m 


Os -t — 


T 


— 


fi js 


Q, 


OONOi 


<o 


oo co ci 


r» 


l— N 


O SO 


t^ 


fN 


t v) a 


i." 


c/5 


V 


c 
o 


I 


Os © m so 


O 


m m «-* 


m 


CN 


o o 


— ■ 


OS 


o o — 




ii 


I 

-fc- 


©'©'©'© 


©' 


©'©"©' 


© 


O 


o o 


o 


O 


o o o 












r^ os — ■* 


ro 


<N — < SO 


so 


<n 


(N fN 


m 


V) 


O O "t 






■e « 








O 


~- fN »-i 


© 


o 


T— 1 ^1 


o 


♦-■ 


•— < »— C »— H 






11 






©'©'©"© 


© 


© o o 


© 


o 


o o 


o 


o 


o o o 






Si JS 

(/J Q 
























on 

CQ 


- 

4< 




c 
c 




oo r- oo so 


r- 


m oc »o 


fN 




— SO 


«n 


t»» 


r- o o 


E 


^ 


ta 


| 


aa-rf 


<N 


Tt- rr m 


m 


■^r 


V) — 


cs 


^* 


fN Tfr *T) 




£ 








© o — «' *+ 


O 


© © © 


© 


o 


O O 


o 


o" 


o o o 






a 




























■ J 


S 1 




os © S 




8S~ 


SD 

o 


(N 

00 




00 


d~> 


o ° S 

Sl; fN os 






05 




























k. 




























V 


























u 


US 

e 
J 


"e 
o 
\E 

'■5 

c 
o 

U 




verage 

igh 

ery High 

ooded 


O 


verage 
igh 
ery High 


fc 


1 


verage 
igh 


•a 

X 

<L> 


0JQ 
■ 
E 

E 

> 


igh 

ery High 
looded 










< I > u. 


< X > 


Q 


►3 


< X > 


< X > E 












£ 














^ 














o 














o 








75 






c 


S3 












Cm 








e 






3 


CO 




Sjj 








"3 








2 






o 

o 


5 
£ 




CO 

o 








O 














Z 




2 








z 








■ 

- 

eg 
I 








2 




















£ 








U 




















u 








C/) 

















330 



-o 
<u 

3 
fi 

♦■» 

c 
o 
o 

I 
I 

00 
I 

m 

CO 

H 

















































5 
































ON 


•-H 


■• JTl 






ii 
































o 


o 


so r- 




0. 


c 
5£ 
































o 


o 


O r«^ 


































o 


o 


o o 




































d 


d 


d d 






































t-« 


tN 


£; £ 






b 
































fN 


VO 


^? <^> 






— 
































VO 


O 


S^ o 






_© 
































«o 


00 


O vo 






un 
































d 


d 


-r d 


.2 


j: 




- 

4. 




































I 


c. 


_£ 


re 






























<n 


00 


vo m 


_u 


I 


a. 


£ 


■c 




























m 


m 


o S 


■3 — 


a 


£ 


■ 

a. 

4J 

a 




























fN 


rn 


Tj- VO 


•LB 


■ 


u 

II 

> 


c 
o 




























fN 


r^ 


fN *~ 


II 


re 
* 


a 

re 

■*- 




























d 


d 


d d 




1 


1 
o 








































« 


■J 






■* 


«o 


«/~l 


o 


*n 


r<^ 


fN 


</-> 


00 


oo 


»o 


o 


V> 


s 


«o 


o m 




■e 


re 






o 


•— < 


o 


>— • 


o 


O 


•— < 


© 


o 


o 


o 


'-" 


"- 1 


'- < 


^ •— i 




1 


1 






© 


© 


© 


o 


© 


© 


© 


d 


d 


d 


d 


d 


d 


d 


d 


d d 




m 


a 






































H 


re 
— 




c 
c 




o 




tN 


so 


t^ 


o 


m 


00 


tn 


m 


so 


m 


r- 


«o 


(N 


o r>- 


V 


•s 


I 


r*> 


SO 


© 


tN 


^H 


m 


«T> 


o 


cs 


oo 


o 


tn 


so 


fN 


*r> 


F> r^ 


a .= 

|0 


re 


o 


© 






© 


d 


© 


d 


1 """' 


d 


d 


^ H 




d 


d 


d d 




o 










































"c. 
£ 

re 


J 


■* 




tN 


SO 
tN 


SO 
tN 


tN 


SO 

o 


tN 

00 




</■> 


00 


m 


Ov 


o 

fN 






SO 
(N 


SO — 
tN tN 




M 










































k 










































■ 










































fl 




■ 




































■ 


re 


X 


s 
o 






U 




-a 






O 




•a 


ej 




•a 


T3 




U 


"5. 


li 


■5 
c 






60 




s 






00 

P 3 




2 


00 

CO 




2 






00 


X 


35 


1 

c 
I 




I 

u 




© 


c 
B 

> 


X! 

op 




b 


o 


c 
6 
> 


op 


b 


u 
> 


x: 

00 


b 

<L> 


O 
O 


o 


u. 

> 


X! ?-. 

00 •-" 




O 








— 


< X > 


Q 


_1 


< X > 


< 


s 


> 


U, 


— 


< K > 






























o 


















] 






^— 
























— 










■o 
o 






53 








33 










3 


















£ 






5 

s 








iv5 
O 

2 










O 

o 








5 
^ 








■ 
o 


































1 








« 








U 


































55 
J 










o 

on 


3 
O 
&0 
























CO 


s 

s 







331 



■a 

<D 

3 
C 

i 

o 



00 

I 

rn 

o 
X> 

H 

































O 

ii 




— 


_ i 


— — 


o 


rl" CN 


CN — 


oo — <n — 


__ 


VO 


H 


— NO 






o 


o 


O On 


m 


no o 


o © 


m o oo — 


o 


CN 


O 


no r- 


a. 


s. 

e 




o 


o 


© o 


o 


— oo 


o o 


m o — vo 


o 


CN 


© 


>n On 






o 


o 


o o 


o 


o ■* 


o o 


o o o m 


o 


O 


O 


T F" 




S 




© 


o 


o © 


© 


© o 


o o 


© © © © 


d 


O 


o 


© o 








m 


tT 


NO rr 


o 


2; ■«■ 


— CN 


n 5 fc S 




<o 


ON 


<n o 




o 




V> 


«n 


00 — 


o 


£ <N 


\C t*~ 


Tf 00 )0 £- 


"• 


.— 




t-~- o 






f-^ 


o 


o r-» 


■xfr 


}Q s& 


o S 


vc oo JN 00 


* 


SO 


00 


— F 




en 




cn 


o\ 


vn o 


on 


^ ^F 


rj- so 


no*"! 


«r> 


oo 


t> 


m o 










o 


o © 


o 


7 © 


o o 


~ o t 9 




o 


© 


o o 


<u X 


k 

4/ 


























.dj 
II Gal 

Dept 


If 


, 


•* 


CN 


o m 


o 


t> NO 


o o 


r^ tj- ^ cn 


t^ 


NO 


00 


On 00 


■S 


no 


ON 


CN SO 


00 


00 00 


5 © 


V> 00 P-- SO 


00 




00 


CN CN 


B 

I J 

55 


c. 


*-* 


cn 


O CN 


«n 


ON CN 


o F 


^f so S o 


CN 


wn 


h« 


r^ «r> 


•V K £ 


0. 

a 


</n 


CN 


cn m 


t^- 


ON O 


M rn 


— < CN — O 


m 


o 


m 


o — 


11 « 


© 


o 


© © 


o 


o © 


O O 


O O © O 


o 


o 


o 


© o 


■o 


e 


























1 


1 


























■ 


•e 




oo 


NO 


•— i >n 


ON 


<r\ rn 


CN SO 


CN O — ON 


rr) 


CN 


<o 


■^r wn 


■o 

1 


ec 




o 


O 


— o 


o 


»— < ^-h 


i— .— i 


— CN CN O 








© o 


1 




© 


o 


© ©' 


©' 


© © 


o o 


O O O O 


O 


o 


o 


o" o 


c/5 


Q 


























■ k * 






o 


CN 


— <* 




•*r m 


00 Os 


t"» rn o r~- 


CN 


NO 


On 


rf — i 


-■St ■ -c 


? 


CN 


Tj" 


no NO 


00 


rn m 


rf \C 


— *r> oo r»- 




CN 


tT 


no r- 




U^ 




©' 


© 


© o 


o 


© © 


O O 


© © o o 


O 


o 


© 


©' ©' 


Jj 




























"a 

E 

I 


J c 

en w 




NO 

o 


CN 

00 


</n 


oo 


w 2 


o n- 

CN Os 


«n 3 vo — 

CN ^ CN CN 


O 


CN 

00 


m 


vn oo 


en 




























k 




























V 




























A 


■ 


























O OS 

SS Li 


B 

_ e 

51 








1) 

00 


■a 

X 


CO 


•SP-o 

u o 
> E 


(U -XT 






00 

9 


X 


J 


o 






Q 


o 


ft •*= P"> 

£.2P £ 

< I > 


< s 


o > •— <u 

J < ffi > 


Q 


o 


> .SP B 

< I > 
















£ 
















_ 












o 




, 












4/ 




























e 




— 








8 




i/5 


^2 










s 




i/5 
o 
2 








1 






o 
2 








e 




























o 




























a 




























■ 

en 


















j? 










Jf 


















O o 










B 
■ 

u 


















5 g 











332 



3 
C 

I 

O 



00 
I 

m 

o 

3 
H 



p 

(Slope = 0) 


O 00 — • — 
00 — i o o 
m in o o 
o <n o o 

© o o © 


o o o o 
o o o o 
o o o o 

©'©©'© 


O O O SO o 

o o o >n o 
o o o o o 
©"©*©' o ©' 


cs © © © 
<n © © o 
© © © © 

© O © © 


a 
e 

53 


-3.2986 
0.5791 
0.6946 
0.9472 


rN r* <— i oo 
r» *o tj- »n 
r-» — i <n en 

00 00 00 00 

O o o © 


Os On 00 r~ © 

<N os m oo so 
© — < K o ■* 

<N "i On tt f* 

— > »-« © «-< ©" 


n n m rt 
<N «n tj- Os 
■* >n rs o 

SO 00 On O 
<N ©' ©' »-» 


(X=Sill Gate 
Water Depth, 

Y=Check 

Station Water 

Depth) 


On <* O tT 

fN \D N t 

os m r» so 

os o — ^r 

O o © o 


Os so C» Os 

SO t OS - 
(N (#1 O (N 
so Os </"> 00 

© o © © 


— — (N Tt ^r 

oo rsi m os «n 

Tf <n 00 SO — 

os r~- oo so os 

©©'©'©' © 


<ri m t~~ «r> 
so oo S r^ 
oo rj- <n oo 

OS NO On Os 
© O © O 


•a c 
u e 

a a 

§ f 

5 £ 


oo oo so — 

fN «— — (N 

o o" o ©' 


* m no os 

o «■* o o 

o o o o 


so "* m oo t^ 

© © <— © o 

©' © © © ©' 


(N <0 © <TN 

(N O « - 

© © © ©' 


Average 
Water 

Depth at 

Check 

Station 

(m) 


so o r^ so 

— " — (N «-> 

O © o © 


O SO tN ^< 
•n r-~ pn r«"> 

O* O — h — i 


rr os rr o m 

t <fl OOfN t 
©' © o — i — ' 


(N O IT! 00 
On — Tj- f«. 

© —>' — ! — ! 


Sample 
Size 
(n) 


n>Ort 

m jU <n os 


(N ^ fs| (N 


)S ro ** 

O £} <*> <ri oo 


<*> 2h <N Os 


Sill Gate 

General Water 

Level 

Condition" 


Average 
High 

Very High 
Flooded 


Low 
Average 
High 
Very High 


Dry 
Low 
Average 
High 
Very High 


Average 
High 

Very High 
Flooded 


B 

X 

e 


* 

o 

■*-» 
3 

o 

o 

2 


«^4 


»3 

o 
Z 


1 
I 

o 
Z 


c 
e 
a 

i 

Jtf 

I 




O 
SO 

H 

























333 




u 










x 










■*-» 










o 










-♦-» 










T3 










C 










O 










a 










c« 










<D 










E 










o 










o 










T5 










C 










B 










** 










M 










X 










■*-» 










a. 










B 










■o 










u 










<u 










■+-J 










«s 










£ 










u 










+- 










cS 










bo 










__ 






























VI 










TJ 










fi 










J4 










C 










3 










B 










(m 










O 










■*-* 










o 




















a 










rt 










c 










o 










n 










B 










DO 










S 










x 










5 










& 




| 








•»- 












C/3 




i 








i 




01 

c 








1 












c 
is 

u 




o 

V 




V"> 00 CS VO 






Q 




r^ oo o tt 




■s 




u 




© © ^ p. ; 




g 




•4-» 




VI VI VI VI 




1 






■* 


X X X X! v© 






rs 


a Q. O. D.^ 




C/5 




u 

1 

a 


©' 


U (U 1> (L> — i 




c 
_o 


u 

00 

c 
as 


VI 

x 


T) T3 "O T3 A 

V V V V ^ 


3 
C 


; 5 




a. 


«r> *0 Os <^i ■{=! 
(N fl OO O g" 


'•fl 


o 


x 


In 


-o 


© © © »-J "B 


c 


o 


■*- 






o 
□ 


1 


a. 
u 

T3 






00 

i 
u 


Q 


00 

c 

i 

a 

o 


c 
o 
U 




low 

average 
high 

very high 
flooded 



334 
The Central basin, represented by check stations in Chesser Prairie, Coffee Bay, 
Gannett Lake, and SCRA behaved independent of the sill's presence in all but the highest 
water levels (Figure 3-19). No significant changes in hydroperiod frequencies in this 
basin occurred with removal of the sill (Figure 3-18). Under low water conditions the 
river floodplain drains more rapidly than this area, which loses water primarily through 
evapotranspiration (see System Sensitivities section). Under average conditions water 
levels reflect those at the sill gate when the sill is in place slightly more than when it is 
removed, but not enough to change hydroperiod frequencies. Under extremely high 
water conditions drainage from this area is delayed, most likely as the head difference 
between this basin and the west basin declines. The slope relationship changes from 
positive to negative under these conditions, and this is exacerbated by the sill's presence. 
This suggests that under these conditions (occurring during 6% of 1980-1993), drainage 
may occur to an alternate basin, possibly towards the St. Marys River basin in the 
southeastern swamp. In "no-sill" model simulations water elevations reached this 
condition during 2% of 1980-1993. The Moonshine Ridge area, in the St. Marys River 
watershed, also demonstrates this shift in relationship with sill area water level 
fluctuations. Drainage of this area under extremely high water level conditions may 
occur more quickly than in the western and southwestern swamp due to the smaller 
volume of water collecting in the watershed (Figure 3-20), although a backwater effect 
may also be occurring in this basin as water accumulates in the St. Marys River. 

Chase and Territory Prairie water levels, when very high, are also negatively 
correlated with those in the sill area when the sill is present (Figure 3-19). The 



335 



37 



< 

I 36 

c 
o 



• 



3 35- 

(0 



I 



34 
37 



i 

< 

£ 36 



1 

LU 

■ 

I 

3 35 

CO 



i 



34 



Sill Area With -Sill 



Billy's Lake With-Sill 



Billy's Lake No-Sill 







Sill Area No-Sill 






Sill Area With-Sill 



SCFSP With-Sill 



SCFSP No-Sill ^:^.'.:r'•r•%? i ,^ : ?Tv.•' , ;' : 

MM* 






t ' 



Sill Area No-Sill 



T 



T 



Low 



Average 
Sill Area Water Level Condition 



High 



Figure 3-19. Comparison of semi-monthly water surface elevations at recorder stations 
and the sill gate area during 1983-1993, in "with-sill" and "no-sill" model simulations. 
Data are arranged to illustrate the change in water levels at the selected station, relative 
to increasing water depth at the sill. 



336 



37 



i 



< 

£ 36 

c 
o 

1 

© 

i 

= 35 

CO 



34 
37 



Sill Area With-Sill 



Craven's Hammock With-Sill 




Craven's Hammock No-Sill 



.\>-^ 






Sill Area No-Sill 



34 




Suwannee River No-Sill (same as sill area elevation) si " Area No-Sill 



Low 



i ~i i r 

Average 
Sill Area Water Level Condition 



High 



Figure 3-19 -continued 



337 



38 



CO 

E 



■2 36 
ui 

1 

1 

o 35 

I 



34 
37 



I 

< 

£ 36 



I 

* 35 



I 

i 



34 




Sapling Prairie With-Sill 



Sapling Prairie No-Sill 



Sill Area With-Sill 



Sill Area No-Sill 



:<*y ••<,-• ; 



Floyd's Prairie With-Sill 




Sill Area With-Sill 






Jt . /_ r>o"w.^V>"'." '•.'•" "' • i : 



Sill Area No-Sill 



Low 



— i 1 1 r 

Average 
Sill Area Water Level Condition 



High 



Figure 3- 19-continued 



338 



39 



<0 38 

S 



| 37 

to 

> 

J 
EO 

1 

8 36 

t 



£ 



I 

| 35 



34 
39 

w 38 

2 

1 

! 37 



f 

I 36 



I 35 



34 



^^WW^^ 



Durdin Prairie No-Sill 




Sill Area With-Sill 



Sill Area No-Sill 



v;.:.^^T 



: f 



^^^^ 



Double Lakes No-Sill 



Sill Area With-SIII 



Lwf& 



Sill Area No-Sill 



; ; ,, ~.».-^.k 



Low 



~i 1 1 r 

Average 
Sill Area Water Level Condition 



High 



Figure 3- 19-continued. 



38 



CO 

S 
< 
E 



37- 



■2 36 



i 

$ 35 
I 



34 
38 



CO 

S 37- 



1 



® 36 



CO 

5 35 
I 



34 



Territory Prairie With-Sill 




Territory Prairie No-Sill 



Sill Area With-Sill 



.. v .~- ■"•-•••'- • • ' 
Sill Area No-Sill 



pt'XV* 






! .*. ..■•■.: ^; 
i ... .•xr- '•*■.":•'•• 



Chase Prairie With-Sill 







Sill Area No-Sill 



339 



Low 



- 1 1 1 r 

Average 
Sill Area Water Level Condition 



High 



Figure 3-19-continued. 



340 



38 



-J 

1 

1 37- 



u 36- 



I 35 

i 



34 
38 - 



co 



37 - 



■S 36 

UJ 

e 

I 

3 
CO 

5 35 

(0 

5 



34 



Chesser Prairie With Sill 




Chesser Prairie No-Sill 



Sill Area With-Sill 



( ..,,,,^>fY^ j: 



Sill Area No-Sill 



Gannett Lake With-Sill 







^.^^^f*^^ ? ! ' ' 



Sill Area No-Sill 



Low 



^ i i r 

Average 
Sill Area Water Level Condition 



High 



Figure 3-19-continued 



341 



38 



CO 

1 » 

E 



! 

•2 36 



* 35 
I 



34 
42 



CO 
S 
< 

E 



41 - 



40 - 



§ 39 



1 



■£ 38 
UJ 

■ 
g 

£ 37 

3 
CO 

5 36 

i 

35 - 



34 



Coffee Bay With-Sill 










Sill Area No-Sill 



.^N^v^^ 



SCRA No-Sill 



Sill Area With-Sill 






Sill Area No-Sill 



Low 



Average 
Sill Area Water Level Condition 



High 



Figure 3- 1 9-continued. 



342 



38 



CO 



< 

£ 37 

c 
o 



1 



© 


o 36 
i 



J 35 



34 



Sapp Prairie With-Sill 




Sapp Prairie No-Sill 



Sill Area With-Sill 



,_„.....,.......•*.--.•••••••'•••"••• 

Sill Area No-Sill 



> 



Sftj/fe^rCT 



-■'■•; % j | ' 



38 - 



CO 

S 

1 37 



UJ 36 



I 3 5 
I 



34 



Cypress Creek With-Sill 



Sill Area With-Sill 




c-siS'V^ 






Sill Area No-Sill 



Low 



~~ I I 1 I 1 - 

Average High 

Sill Area Water Level Condition 



Figure 3- 19-continued 



343 



37 



< 
£ 36 



i 



3 35 

to 



1 



34 



Sweetwater Creek With-Sill 




. ■ c •...•*••'.•;.■■• ' " 
Sill Area No-Sill 






CO 36 - 

S 

1 



UJ 

e 

§ 35 

t 

3 
CO 



I 



34 



Sill Area With-Sill 




Sill Area No-Sill 



Transect 60 No-Sill 



Low 



Average 
Sill Area Water Level Condition 



High 



Figure 3-1 9~continued 



344 



38 



CO 
S 
< 
E 

I 



37 



■S 36 



I 



t 

3 
CO 

£ 35 

I 



34 



Moonshine Ridge With-Sill 







Sill Area No-Sill 



Low 



T 



T 



Average 
Sill Area Water Level Condition 



High 



Figure 3- 19 -continued. 

















345 










250 - 
























St. Marys River-Moniac 




200- 




i 
• 




- - - - Suwannee River-Fargo 












^•^ 




i 






CO 




» 






E 










o. 




■ 










s 






© 




s 




| 




-4-> 




*l 




1 




CO 

a 150 - 




•i 
•• 




! $ 




2 




'1 




: ; J 




o 

UL 




1 

; f. 




: { $ 




>l 




\ 




\ 




3 

• 

|100- 

ffi 
o 

O) 




?! 

i: i: 


i 
• 

! 

1 

% 


It ' \ 

■ !. 

■ : J ■■ ' 

' s •■ •• "' 




2 

CO 


i. 


■ • • 
i ' » ' 


. "i "i ■'. ■' *•'■ 




> 




■ ■ 


1" ' • ' ■ .1 * . • 1. • »■#* 




< 50 - 


i ■ 
• • 


■ ^ 
; i . i .; 


• • ■• ' ■ :• 2 ffl' 






■ ■ 


; ■ j ; \{\ 


' ' It ' 1 I 1 , 1, 






• i 


•■ ■ Jpifi 


■■J" ■ • ! i! '! ■ •• ' ' '. " i 






i 


& Mw- 


^fLfi^lLi : J il&L> 




- 


<»^ 


hh^nA) UrU, 


iOdWlW l^lw^a#AMi_L^yw^ 




i 


1 1 1 1 


i i i i i i i i i i 






1980 


1982 1984 


1986 1988 1990 1992 1994 










Year 







Figure 3-20. Estimated average biweekly flow rates at Suwannee River (Fargo) and St. 
Marys River (Moniac) during 1980-1993. 



346 
impoundment of the sill and the natural berms in the river floodplain delay de- watering, 
permitting water to backup into this region during these extreme flood events (Figure 3- 
21). As in the Cypress Creek drainage and Chesser Prairie-Coffee Bay region, changes in 
hydraulic head differentials are most likely driving this varying relationship. During 
average and low water levels, this correlated relationship is absent, regardless of the sill's 
presence, although the removal of the sill may slightly alter the frequencies of 
hydroperiods in the deepest classes (Figure 3-15). 

Hydroperiod frequencies do not change in the Northeast basin with removal of 
the sill (Figure 3-18). During average and low water level conditions, water levels are 
more closely correlated with those at the sill when the sill is present than when it is 
removed, but this does not effect flooding depth group frequencies and durations (Figure 
3-19). During periods of extreme high water, the correlation declines, regardless of the 
sill's presence. 

Frequencies of hydroperiod depth groups significantly decrease 1-2 classes in the 
southwestern creeks (Sweetwater and Cypress) and Sapp Prairie with sill removal (Figure 
3-18), although there is much unaccounted variability in creek water levels during 
average and low levels (Table 3-6). In Cypress and Sweetwater Creeks water levels 
decrease with increasing levels at the sill when extreme highs occur, indicating a switch 
in the hydraulic head (Figure 3-19). In extreme high water events the sill restricts de- 
watering of the western swamp while the creeks continue to accumulate and drain water 
from their watersheds. As the creek level falls, the creek-river hydraulic head reverses 
and the creek outflow is restricted by river flow. When the sill is removed this slope 



347 



1 



two c*5 $ co 8 "i§ j-f 
pfiat ■ ■ • « |5 




I 

1 



CO 

u 

(L> 

J2 
g 

V4-1 

V 

-c 
O 
u 



«3 

"a. 

T3 

o 
o 

a 

1- 

s 

5 

a> 
<u 

c 
c 

3 

co 



C 
o 



5. 

& 

a 

o 



C u 
o > 

a 2 

3 £ 



1 s 

m co 



I: 



348 
relationship becomes nonsignificant, suggesting that without the sill, de-watering of the 

swamp through the river supersedes that from the creeks at high water levels, resulting in 
delayed drainage of the southwestern basin (Table 3-8). A berm in the Suwannee River 
southwest of Cypress Creek slows the river drainage even in the sill's absence, creating a 
backwater impoundment in the Cypress Creek basin out of the recorder's range. Levels 
in Sapp Prairie do not reflect the switch in hydraulic head that occurs in the Cypress 
Creek basin, although a slight decline in water surface elevation occurs in the "no-sill" 
simulation (Figure 3-19). 
System Sensitivities 

The model was also manipulated to increase the volume of water impounded by 
the sill; all water exiting the swamp in the Suwannee River was retained in the swamp by 
setting the "Suwannee River outflow coefficient" to for each decade interval (Figure 3- 
4). This approximated the maximum possible impounded volume, and indicated the 
largest area that would have been affected with the current sill configuration, historic 
inflow, precipitation and evaporation, and no river outflow. The area bordered by 
Craven's Hammock, Billy's Lake, and Transect 60 increased 1-1.5 m due to the reduced 
outflow; Floyd's Prairie and Coffee Bay increased roughly 0.40 m and no change 
occurred at Sapling to Chase Prairies and beyond (Figure 3-22). These increased water 
depths were similarly observed during all decades. Inundation intervals were also 
prolonged where the increased depths occurred, with most intervals at depths >1.0 m 
(Figure 3-18). Water levels decreased in the Cypress Creek and Sapp Prairie regions 



349 



20 








20 



40 Kilometers 



A/ sin 

/S/ Refuge Boundary 
/\y Islands 

■MB Islands \y 

/\/ Contour Intervals (m) 




Figure 3-22. Inverse-distance-weighted, contoured estimates of increases in average 
semi-monthly water surface elevations (m) at recording stations, attributed to Suwannee 
River outflow retained in the swamp during 1980-1993 model simulations. 



350 
with this manipulation during 1941-1949 and 1960-1993 (Figure 3-18). Outflow in these 

areas may have accelerated as the hydraulic head between the Suwannee River 
floodplain and the Cypress Creek watershed increased with reduced river flow volumes. 
However, low water levels in the Cypress Creek drainage during 1950-1959 reduced this 
hydraulic head difference between Sapp Prairie, Cypress Creek and the Suwannee River 
floodplain, and hydroperiod group frequencies were not different from those occurring 
with normal river outflow. 

Manipulations of evapotranspiration (ET) rates to examine potential effects of 
vegetation change on swamp water levels indicate that regional differences exist in the 
importance of this process to the swamp hydrology. ET was decreased during 1980-1993 
to 75% and increased to 150% of estimated volumes with river flow rates set at best 
"with-sill" model conditions (Suwannee River outflow coefficient^). 15). Changes in ET 
volumes had greatest effect in the eastern and central swamp; water surface elevations 
dropped below "no-sill" levels at 150% ET and exceeded "no outflow" levels at 75% ET. 
Changes in the water surface elevations in the western swamp were less extreme, 
approaching "no-sill" levels at 150% ET and "no outflow" levels at 75% ET (Figure 3- 
23). These responses contrast those resulting from outflow volume manipulations. 
Western swamp water levels fluctuate with changes in outflow proportions, whereas the 
eastern and central regions remain relatively stable. Regional differences in vegetation 
distributions and topographic relief drive these responses. The greater topographic 
gradient in the river floodplain emphasizes changes in outflow volumes in the western 
swamp, and the prevalence of open water, aquatic and herbaceous prairie, and the low 



351 



2.0 



1.5 - 



f 10- 



E, 

5 

a 

& 



£ 



0.5 



0.0 - 



2.0 



1.5 - 



£ 1.0 
I 



I 



0.5 - 



0.0 - 



Best Model 
75% ET 
150%ET 



Billy's Lake 




i — i — i — i — i — r 



i — i — i — i — r 



Best Model 
75% ET 



SCFSP 




— I 1 1 1 1 1 1 I I I I I I I 1 

1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-23. Manipulations of estimated evapotranspiration rates and responses of the 
model at recorder stations during 1980-1993. 



352 



3.5 -i 








Moonshine Ridge 






Best Model 


3.0 - 




75% ET 

150% ET 


• 




2.5 - 






p" 








a 
I 


rv 






f 1.5- 

a 

5 1.0 - 


* * 


i\\ & 


L t^ * A 


0.5 - 


Mhh 


w> 


khhlwi 


0.0 - 


vNu V) 


v\jv 


' VvvJu wvw 






I I I I 


I I 


I I I I i I I 



3.5 



3.0 - 
2.5 



J 2.0 - 
? 1.5- 



i 



1.0 



0.5 - 



0.0 - 



Best Model 
75% ET 
150%ET 



Sapp Prairie 






v 




1980 1982 



— I 1 1 — 

1984 1986 



i — r 



i r 



1988 1990 1992 1994 



Year 



Figure 3-23-continued. 



353 



t.o 












Double Lakes 






•»__» W--J-I 


2.0 - 






75% ET 
— 150%ET 


l * 1 

1 




t 

■;»■ 


? 1.5- 

i 

1 






fa ••' '» 


0.5 - 












0.0 - 
















I 


I I I 


1 


1 1 1 


i i i 1 1 1 1 



2.5 



2.0 - 



E 1.5- 



% 1 ° 
f 

I 

0.5 - 



0.0 - 



Best Model 

75% ET 

— 150% ET 



Durdin Prairie 



/ ■- 



af v 



:! \» r> 



■ re- •: 



M '•; 




I I I 1 1 

1980 1982 1984 



1986 1988 
Year 



I I I 1 1 — 

1990 1992 1994 



Figure 3-23-continued 



354 



3.5 -i 














Chesser Prairie 


3.0 - 


Desi rvioaei 
75% ET 
150%ET 


2.5 - 




Ik 


/ ' V 


f,0- 

a 
i 

f 1.5- 

* 1.0- 


i 

.\ / V 

■ V 

.* X 

MwvA A AwM A MJw 


0.5 - 


^W^WWm 


0.0 - 


W V \J ^ 






I I I I 


i I i i i i i i i i i 





a.u - 














Gannett Lake 




dgsi Mooei 




2.5 - 




75% ET 

150%ET 






2.0 - 
1.5 - 






.'V. ••' \; 'vv 1 . . ; • 


? 
£ 

a 




;•. ."Vv 

r 


; ' \; v •', A At » 


1 

1 


1.0 - 


.r v r\ ? 

-v.-v \ J V.: 




0.5 - 


'WWwkAM 




0.0 - 


V v v M il vMyif ^y 






1 1 1 1 


1 1 1 1 1 1 1 1 1 1 



1980 1982 1984 1986 1988 1990 1992 1994 

Year 



Figure 3-23--continued. 



355 



3.5 
3.0 
2.5 



J 2.0 

I 

* 1.5 

1 



I 



1.0 
0.5 
0.0 

2.0 



1.5 - 



5 1.0 



0.5 - 



0.0 - 



E 

f 

I 
i 



- Best Model 
75% ET 

- 150%ET 






"\ A 



Coffee Bay 










i r 



Best Model 
75% ET 
150%ET 



SCRA 



?**. s 




i i i i n 

1980 1982 1984 



1986 1988 
Year 



1990 



— i — i — i — 

1992 1994 



Figure 3-23-continued. 



356 



3.0 



2.5 - 



2.0 - 



a 1.5 

I 

a 

1 

« 1.0 



0.5 



0.0 - 



2.0 



1.5 - 

? 

£ 1.0 



I 



0.5 



0.0 - 



Best Model 
75% ET 
150%ET 



Territory Prairie 



- * i > 



.-*- 



~a/v 







t — i — r 



Best Model 

75% ET 

— 150%ET 



Chase Prairie 




i r 



i — r 



i — r 



1980 1982 1984 



1986 1988 
Year 



I I I 
1990 1992 



1994 



Figure 3-23-continued 



357 



2.5 

2.0 - 

? 1.5- 

f 

! i.o- 



0.5 - 



0.0 - 



2.0 



1.5 



§ 1.0- 
I 



i 



0.5 - 



0.0 - 



Best Model 

75% ET 



150% ET 



Floyd's Prairie 




i — i — i — i — i — i — i — r 



Best Model 
75% ET 
150%ET 



Sapling Prairie ■-, 



r. ■ - 
V ' va? w 






!> V 




1 1 1 

1980 1982 



1 1 — 

1984 



i — i — i — i — i — i — i — r 



1986 1988 
Year 



1990 1992 1994 



Fig\re 3-23"CQntinu(?d, 



358 



4.5 
4.0 - 
3.5 - 
3.0 - 



5 2.5 

a 
i 

° 2.0 - 

] 

5 1.5- 



1.0 - 
0.5 
0.0 - 



Best Model 

75% ET 
150%ET 



Cypress Creek 



.-. <>-J 






*W 



: r V« 






,v'-v 




Av/UL 



i i i i i — i — i — r 





n.v 
3.5 - 














Sweetwater Creek 




' ' D6SI MOuGI 

75% ET 




3.0 - 
2.5 - 
2.0 -i 
1.5 - 
1.0 - 




150%ET 




? 

5 

a 
o 

Q 
l. 

£ 

(0 

5 




A 1 

v Aw 






0.5 - 


'W^MfcftPMjd^ 




0.0 - 


v\j v vwy v wx^[\$} v w \ji 






1 1 1 1 


1 1 1 1 1 



1980 1982 



1988 1990 1992 1994 



Year 



Figure 3-23-continued 



359 



2.0 



1.5 - 



f 10- 

I 
O 



I 



0.5 - 



0.0 - 



2.0 



1.5 - 



f 10- 

I 
Q 



I 



0.5 - 



0.0 - 



Best Model 
75% ET 
150%ET 



Craven's Hammock 




i r 



Best Model 
75% ET 



- 150% ET 



Suwannee River 




1980 1982 1984 



I I I I I I I I I — 

1986 1988 1990 1992 1994 

Year 



Figure 3-23-continued. 



360 



2.5 



2.0 



? 1.5- 
l 



® 
o 

© 

I 



1.0 



0.5 - 



0.0 - 



2.5 



2.0 - 

? 1.5- 

I 

I 
O 



i 



1.0 - 



0.5 - 



0.0 - 



— Best Model 

75% ET 

— 150% ET 



Sill Gate (south) 




i r 



i — i — r 



Best Model 
75% ET 
150%ET 



Transect 60 




"~ I 1 1 — 

1980 1982 



— I 1 1 — 

1984 1986 



1988 



— I 1 1 — 

1990 1992 



1994 



Year 



Figure 3-23--continued. 



361 
topographic gradient are probably responsible for the importance of ET in the water 
budget in the remainder of the swamp. 
Wildfire Occurrence 

The Suwannee River sill was constructed to eliminate or arrest wildfires in the 
Okefenokee Swamp (Chapter 742, Public Law 81-810, 70 Statute 668). During 1960- 
1993 wildfires continued to burn throughout the swamp and in the area impounded by the 
sill (see Chapter 5). Burned area decreased after sill construction, although this decrease 
was probably not due to the sill since water levels were low or at drought levels when 
most of the fires were ignited, and the fires occurred outside of the low- water and 
drought impoundment areas (Figure 3-24). The decrease was more likely due to fire 
suppression efforts, and the absence of severe drought during 1960-1993 (see Chapter 5). 
More fires were reported in the Okefenokee Swamp during the with-sill period (151) 
than during the century prior (98) to its construction. Since 1855, 37 fires were reported 
in the area affected by the sill impoundment; 18 of these fires were prior to sill 
construction, and 1 1 were in the Cypress Creek watershed (see Chapter 5). All of the 
fires occurring after 1960 were extinguished by fire suppression efforts or precipitation; 
none were arrested by the sill impoundment. Water levels were at low or drought levels 
when 16 of the "with-sill" fires occurred. These fires were burning outside the region 
impounded at low water levels, and probably would have been ignited and burned if the 
sill had not been present. They would have been arrested by the sill only if they burned 
into the low-water impoundment area. 



362 




s 






a 



> 

2 

u 

c 
c 

3 
v 

>> 
X) 

i 

o 



a: 



363 
Fire exclusion throughout the swamp will never be achieved with the present sill 
because of its limited impact area (15% of the swamp at high water levels) and 
seasonality of impoundment. The sill increases water levels at high and low water 
conditions in the area encompassed by Transect 60 to Craven's Hammock to Floyd's 
Prairie to SCFSP (Figure 3-9). Its removal reduces the high water levels in an area 
slightly larger than its low water level impact area (Figure 3-9). Only a slight change in 
high and low water levels occurs in the Floyd's Prairie to Sapling Prairie region, and no 
change occurs east of Bugaboo Island with its removal. Most of the wildfires (25) in the 
sill and Cypress Creek areas have ignited during June-October, when lightning strikes are 
most common (Chapter 5) and water levels rapidly decline in the absence of precipitation 
(Chapter 2). Greatest impoundment occurs at high water, usually during winter months, 
when thunderstorms and lightning activity are infrequent, and water level accumulation 
occurs at reduced levels of evapotranspiration. Even if all river outflow were captured 
by the sill, the impoundment would not increase in area (Figure 3-22), although depths 
would increase in most of the current impact area (Figure 3-21). This greater 
impoundment volume would reverse the hydraulic head at the Cypress Creek-Suwannee 
River junction, causing the creek watershed depths to decrease (Figure 3-14). The 
Cypress Creek to Sapp Prairie areas have experienced increases in high water levels 
since the sill's construction, although it is not certain that this is directly attributed to the 
sill's impoundment. Areas between the Cypress Creek watershed and the sill are not 
affected by the impoundment, suggesting that the cause of this increase is independent of 



364 
the sill, and actually attributed to natural impounding occurring in the river floodplain 
southwest of the swamp (Figure 3-15). 
Vegetation Change 

Types of vegetation changes occurring in the sill impoundment area and Cypress 
Creek watershed mirror those in the remainder of the swamp, although change rates 
differ (Table 3-9). Wet forest initially increased in the river floodplain impact area 
during 1952-1977, and was persistent during the next 13 years, whereas shrub, prairie, 
and upland pine coverages were nearly halved during 1952-1990 (Table 3-10). These 
changes occurred at rates slower than the surrounding swamp during 1952-1977, and 
then greater than the surrounding swamp during 1977-1990 (Table 3-11). Shrubs flooded 
during the initial impoundment did not survive unless located on elevated surfaces. The 
apparent increase in proportion of forested area was probably due to this decline in shrub 
coverage. Recruitment of trees and shrubs has been eliminated during the extended 
flooding; only periods of drought provide exposed surfaces for germination, and survival 
of seedlings is jeopardized by flooding occurring before sufficient stature to survive 
impoundment is achieved. As in the remainder of the swamp the impounded area is 
advancing in successional sequence in the absence of severe fire (see Chapter 4). 

Cypress Creek watershed vegetation has converted from prairie, shrub, and scrub 
composition to shrub, bay-shrub, cypress-gum-shrub, and other wet forest-shrub 
associations. Most of the prairie coverage in the watershed was eliminated during 1952- 
1990 (Table 3-10). In contrast to the remainder of the swamp, most of this change 



365 



m 

On 
On 

i 

00 
00 

'C 

3 

"O 

T3 
1) 

E 

3 

C 
ca 



u 
X 

X 

-o 

u 

•** 
o 

I 

CO 
CM 

CO 

<L> 

i- 
« 



00 

c 



3 

o 
u 
o 

u 
00 

c 
cs 

X 
o 

C 
O 

I 

a. 

OJ 

> 



ON 






I 

1 i 

o 1 




































E 


5 

£_ 








■t 

rf 






fN 






o 












*-■§ 


< 








r^ 






— 






o 












S 






































^ 


































1 | 








00 




r-- 

ON 






«n 






NO 




r*« 








■ 

< 






ON 








ON 






r*N. 












e "o 




■3 
































c 4> 




c< 
































- 4> 


< 


E 




Tf 


d 


© 


ON 


*© 


■* 




NO 


o 




NO 






NO 


© < 


I 




SO 


o 
m 


NO 


cn 


ci 


I s -' 

On 


o 


H 


1° 


O 


fN O 


o o 


o 


5 


6- e 




a 
































e v 




































j S 




£ 




— 


On 


m 


o 


On 


O 


<n 




ON O 




<ri rj- 




r- 




© < 


« 




O 


On 


© 


On 


ON 


On 


NO 


o 


o — i 


K 


iri ~ 


<ri 


o 


£ 


1 

as 




oo 


^- 


NO 


CN 


NO 


(N 




vo m 




m 




a. © 


< 


































| 






CO 










t/5 
























Xl 










x 




















c 






5 

x 










1 




















I 






oo 




■*■» 






00 






4-* 


















1 




C/5 






1 






fa 






co 








5 
& 






E 

3 


en 

X 


o 






E 

3 


C/3 

X 


C 


U 

E 
O 






»5 

1) 

E 




c 




> 






o 

1 

n 
n 


3 

j 

OO 


C 


en 


ej 


o 

1 

c/i 

CO 


3 

i 

00 


C 




0« 




>. C3 

U 00 

1 | 


'3 

>— 

| 










2. 


i 


_ea 


X 


•c 


(L> 


1 


_ea 


_2 .o 


•e -2 


X X 


X 


c 










U CQ 




3 
C 
X 
00 


'3 

Du 


O. 
>> 

o 


•2 
CQ 




> 00 


O Xl 

oo oo 


E E 
x J3. 
oo M 


3 

t- 

X 

OO 


cct 

a 


■ 
e 








00 

c 










00 

c 


















S 


Q 






'Bo 

00 










"Ob 
00 


















X, 


c 






_o 










o 




















■ 
1 






1 




(N 












(N 




r>- 








> 






p 




<r> 






u 






m 




t^ 














E 




ON 






u- 






On 




On 
















Oh 




"~" 






0- 










1-H 












































"m 




































t 






■5 






























4* 






J 1 * 






























■S 






Ik 












k* 


















I 






T 












o 


















i 






l 












X 

E 


















3 

oa 

■o 

§ 

at 
t 

< 






^ 
| 


On 










<L> 

u 

Q 






















1 


CQ 

2 










i 






















X 










«n 






















(J 


6 










ON 




















M 






1 


IX, 










— 


in 
















■ 






p. 


VN 










rt 


m 
















* 






5 


00 










2 


3n 

















366 



I 

Sill 



SO 
© 



C - R 
_ = s- 

"I 



Os 



SO 






so 



<N 



= -a — 

o v r 

f J f I 

2 c «- s 

£■8 5 



Os 



ir, 



2 OS 






m so 
o o o r- 



o o o o o 






J 1 1 *S 

II S E 



— — V© 

(N © t--' 

m -— m 



S so 



o o 



IT) 


u-> 


o r* 


rr 


00 <N 


On 

sO 


o 


SO CN 


r>» 


00 — 



© g © o © © 



p 

e 

c 
•o 
« 
— 

I 



= V 

R C 

V B. 



T3 

<D 

3 
_C 

c 
o 
o 

OS 
I 

1 

H 



R 

E 

■ 



ea 

R 

R 

1 



CO 

X 



5-E 



X 



£ 

3 

o 



T3 

c 
to 



00 
1> 



O 
rt 
£ 



x 



oo 






X 

3 
C 

X oo 
00 X) 

« jS 



00 

E 



u 

• - cu 

4 * TD 

D 1) c 
00 <2 « 
•S C — 



o s 

co CO 

6 h 
a. 2 

U CQ 



CO 

<U 

i— 
O 

Uh 

T3 

C 
C3 



f 5 CJ 



co 

X) 

2 



u 



pd 



oo a, 



v 

■c 

| 

a. 

■ 

X) 

5 

X! X 
00 00 



co 

X 

3 



co 
X X 

I 2 
X O 

00 00 



•§ 

I— 

X 

oo 



co 

u .2 



R'l 



co 

cquo. 



X 

2 



X 

5 
s 

oo oo 



X X 

2 2 

X X 

00 00 



00 



o 

Os 
Os 



i s 



Os 



h 

Os 



OS 

I 

P o 

o £ 

z2 



367 



T3 

<D 

3 

c 
o 



ON 

I 

co 

X) 

CO 
H 



•a 




































Total 

nburne 

Area 


"re 


© 










"3- 

ON 

-3- 












co 

ro 






CN 

CN 

CO 




3 




































^^ 




































Total 

Burned 

Area (ha 




CN 










cn 

f— 1 












CO 

1 






O^ 
ON 




rtion 
ected 
ea 


1 


ON 


NO 


NO 


00 


CN 


<n 


00 


NO 




o 




oo 


p 


TT 


NO NO 


o 


9 < < 


I 


On 


<N 

m 


00 

co 


r-' 


vb 


cn 

co 


oci 
co 


I--' 


NO' 


© 




CN 

CN 


— 


s 


oo r-' 

CO CO 


K 


c e 


a 


































S 7 ~ 




































•i 1 5 
f Sj 5 


1 

E 


© 


co 


© 






r-- 


NO 


NO 




»/-> 




— 


00 




On O 




2= « 


<ri 


»— • 


vd 


o 


o 


NO 


</">' 


iri 


On 


SO 




ro 


CN 


o 


— i 00 


o 


3 


•<1- 


*» 






co 


CN 


•— ' 




— 


r^ 




<«■ ■* 




£%< 














































en 

(U 








J 






CO 






















































































































H 

e 




u 

c 


(A 

X) 

3 

J 






_1 

t- 
1 


CO 

I 

X 






« 

CO 


c 




00 
en 

U 

E 


CO 

X 

2 

X 








© 

•■c 
1 




"O 

c 
J2 


00 
1 






00 
1 






T3 

C 




oo 

1 

B 


CO 


CO 




DX 




^ 


3 






E 


3 






I 


■*-» 




u 


3 


X 


O 




j 






co 


X) 

2 




(U 
1 


a 


X 

5 




u_ 

1 






1 


a 
■ 

to 


2 

JS 










"O 

S 


CO 


•C 


CO 


CO 


«« 


X 


CO 


V] 


T3 




PQ 


co 


oo 


C co 


•2 






P 


oo 


X 


V 


B 


GO 


X 


QJ 




1 


<L> 


■ 


cS XI 






a 
U 


>> 

« 

CQ 


3 
X 


00 

■a 

u 
00 


U CQ 


00 


00 

oo 


4> 

X 

2 




a Im *■* 

Bag 

3 >TC 

O U CQ 


^ 00 


'B 

'3 

Cu 


e -— 


























00 

c 










5 


























1 










?. °- 


























o 














o 

ON 

On 










1 


















CN 

ON 








»-N 










•"* 
















t-N 




73 




































t 




































B 




































** 




































e 
























« 












w 
























s 












E 














CO 










c 


CN 










3 

ea 














ON 

ON 










J 


ON 










1 
■ 














1 

o 












b 










! 














On 










T 


C 






















On 










1 


X 












B 
■1 














!• 










1 
























s 










as 












•< 














cn 










9 


00 











368 



■o 
B 

i ■ i » 

H ■£ < ° 



<n 

no 



00 

en 



00 



5 



] I * 

© - 55 

Hit 



■"3- 

On 
CN 
CN 



CN 
CN 
CN 
CN 

CN 



cn 

CN 

CN 



I 

CN 
CN 



c 1 -o 

o ■ 5; 

£ 5 1 

ft. © B 



On 



On 

On O O 



CN 


On O r- 


CN 
CN 


— < O o< On 
■^ CN — < 



o o o 



o 



en 

CN 

NO 



CN m 

no «n 



rr en © 

on «ri on o o 

CN CN CN 



1 "O « 

j 1 1 1 

il i E 



00 

00 



O — 



• 9i _ 

S •— NO ° 



On NO o 

on cn en O 

m ^ - 



l> NO 



l> — ON NO 



jg2(b«* 



O On On 00 

>/S cn <— <n 

CN CN ^< CN 



CO 

X> 



it 

a 

£ 

e 
© 
•G 

55 



55 — 

S i 



•o 

CU 

C 

o 
o 

On 

en 

u 

1 



65 

E 



E 

s 
OQ 

■o 
e 
as 



•9 
& 



X co 




























CO X) 


























X 


CO g 

CO _j= 


























2 

X 


Bay-Cypr 
ss-Gum-S 
Shrubs 




c 
-o 


to 

cU 

— 

o 

T3 

C 


CO 




CU 

C 

£ 

T3 


>> 

03 




Cypress 
Mixed Cypress 


cu 

>- 
a. 


'C 

'<3 
£ 


co 

CO 

CU 
E 

a. 
U 


Ily-Bay 

ss-Gum-S 

tirub 


g K-E c? a 

DUCDfflD 


■♦-* 
cu 


X) 

2 

x 

CO 


'C 

'3 


C 

a. 
D 


X) 

2 

X 
00 


X 

2 

o 

00 


3 

cr 
< 


X 

3 

s— 

X 

00 


X 

2 

X 
00 


Loblo 
Cypre 
Bay-S 



Jg s 

5 3 

x J 

00 t- 

i cu 

co — 

CO CS 

I £ 

>» co 

u E 

i cu 

O oo 



00 

c 

'5b 
ao 

o 

• 

cu 



CN 

On 



ON 



O 

8 



<u 

X 

E 
cu 
u 

CU 

Q 

CN 

On 

X 



22 



369 



■ta 

3 
C 

a 
o 

o 

I 
I 

as 

o 
X> 

« 



Total 

Unbumed 

Area 

(ha) 


VO On n© 
— in ~ 


Total 

Burned 

Area (ha) 


16454 

15734 
16454 


Proportion 
of Affected 

Area 
Unbumed 


oo <n so so oo ^ "* oo «r» 
S <N © © rn — rr" O _• O O O H O • O • 

00— (N N ifl *■> vo 1 ^^ 


Proportion 

of Affected 

Area that 

Burned 


Pith o. "0, °°. «*> tj- m Tf o oo oo so — _ 
§ J5 2 o< $! £ * 2 f* o" K oo r» so so vo ° 


s 
a. 

■ 

e 
•■c 
5 

J 


Cypress-Gum-Shrubs 

Gum-Bay-Cypress-Shrubs 

Briar-Shrubs 

Bays 

Wetland Forest 

Shrubs 

Prairie 

Shrub-Bay 

Shrub-Prairie 

Cypress 

Cypress-Shrub-Prairie 

Scrub 

Aquatic Prairie 

Herbaceous Prairie 

Mixed Cypress 

Bays 

Shrub-Cypress 


Vegetation 
Map Date 


Pre-logging 

1952 
1977 


73 

E 

i 

9 
CD 

1 

■ 

•< 
1 

w 


r^ 

OS 

ka 

u 

X) 

o 
o 

o 

1 

fN 
<r> 

On 

X! 
U 

9 

s 



g 

3 

c 
o 
o 

ON 
I 

u 

•s 



Total 
nburned 
Area 

























no 


</-> 








■>*■ 








« 


^■1 


F— H 








^■M 








£_ 


o 


O 








ON 








3 












































TJ «■ 






















otal 
rne 

a(h 




<N 










CN 








Hit 






















B ~ 


T3 




















oportio 
AfTecte 
Area 




^ ^ °°. *©. VO 00 


ON 


NO 


*> 00 


CN 


m 


oo 


v© 


_ o 


= 

= 
a 


* £ N "* K «ri 

i-H r«^ —< i— i l "* ' 


On' 


<N 


8* 


NO 


(N 


00 


S 


«2 


— e 




















C T5 *. 






















oportio 
AfTecte 
rea tha 


= 

SO 


© 

g O O O o o 


© 


en 


s- 


o 








On' NO 


> <- < 
a. © ^ 
































CO 








V) 














<D 








<u 














^J 








^; 
























« 












-J 








J 


a 






U 


x> 




t- 


X 






u W 






c 


2 

x 






2 

X 








e 






£ 




1 

E 

1 






C3 0- 

L 8 
c — 

6 5 


I 

••e 
■ 

■ 

1 




e/5 

— E -= 

Ik s fr 


c 


00 
1 

O 

Vi 


X) 

2 


00 
1 

E 

3 

5 


X 

2 








■ ■ it 


T3 


C/3 


X to 


1/-, 


</i 


X 


V) 


■5 .a 






X) £ X) X3£ X) 


£ 


oo x 


u 


U 


00 


X 






2 2 2 2 2 5 


J— 


^2 


0X) 


Q. 


>-. 


2 






o x x x ox: 


ii 


5 •*= 


!L> 


>-. 


w 


X 


00 ^ 






OO OO OO 00 OO OO 


u 


CQ oo 


oo 


U 


CQ 


oo 


= o 






















* •£ 






















•c * 






















te Q 












































2° .2 




r-» 


o 








o 








>S 




r^ 


ON 








ON 










o 


ON 








ON 












^ 


"-* 








' — ' 






























"e! 


























>> 


















M 




A 


















1 

9 
CO 




2 




















i 










On 
On 








■o 




t-» 










»-H 








c 
1 
■ 




r^ 

On 










1 

o 








< 

B 














On 










X) 

E 






On 

i 












> ° 
o & 

z2 










2 








< 












CN 




















""" ' 









371 



_ Vi 

Si* 

C B i> 

« c Sj 

Q..2 u 

o 2 * 



o 

x: 



fflu 

> I 

in C 

dJ 2 

E V -a 

1/3 • 5 .s 

u £ o 

■3 •& £ 

9 12 ■ 

*s So 
o ^ r 

3 .2 «J 

O CO u 

©31 



ON 



1) 

— 

o 

u 

r2 
1 

o 



'55 
o 

00 

o 

ja 
o 
es 

o 



c 
o 

1 

00 

> 
o 



13 

c 



4> 

a 



* 6 a 



<N £ O 
3 * | 

00 £ 2 

cUE 

1/5 o 

K •- 

P-E 



3 
T3 



Rt 
173 

<u 

3 
> 

T3 

u 
O 



c 

"S u «n 

•s »1 

« 00 

a a 

CO Q. 

«* Q. 



c 
'53 



OS 5 " 

1 S. E 

E £ C 

u = i 

2 E .£ 



m 



11 



2 € 

o •- 

•S "P 

o, u 

CN 10 

8 H 

2 8 

E 5 

2^ 
da *- 

•o » 

D g 

E J 

20 

.9 < 



e 


<i> 


e 




'-~ 
u 
o 

a. 







e 


■o 


< 


ON 


e 


V3 


mm 




- 


e 


ms 




a. 


M 


on 





u o 



a 

ON 



.2 £ •- 

£ = 

a. bo 



e v s 
•2 j f 

5.3 < 2 

P 2 SB 

S * cc 



.2 *- _ 
8.1 < * 



00 

© 

m 



no on __ 

^ N ^ 



m 

ON 



m rn 



CN 


■<t 


r— * 


On 


On 

NO 


00 


00 


m 

On 






On 

On' 
On 



rf 


>r> 


00 
ON 


NO 
ON 



m 


NO 


r- 


CN 

m 


CM 


CN 



On 



- 



2 ~s 



wn — 

CN CN 



r^ 


Tf 


m 


— 


NO 


00 


wn 


m 

oo 



es 



O 

d 

o 



</~i 

ON 



ON 

r-' 

On 



p = 


~ 




























- 


K 




























.2 j= 
































o 


«N 




o 


ON 


NO 


* 












o 


CN 


8.S 


< 


ON 




CN 


CN 


(N 


r^i 


o 




o 






rn 




o «« 


B3 




























2- — 


«3 




























5 .= 

o — 


J| 




























Proporti 
Outside 


o 


fN 

•r, 

ON 




o 

NO 

r- 




TT 

wS 
r- 


NO 
NO 

oo 


o 

d 
o 




o 
©* 

o 






O 

r-" 

On 


00 

00 
On 


■ 

H 

05 

u 

e 






1 


E 
o 

u. 






a 


e 

3 




i- 


1 




! 

o 

■Q 
B 

£ 
o> 




95 
V 

> 






5 


•a 

B 

4/ 


on 

S 

E 
ja 
65 


.2 
"S 


•a 
a 

iS 
*iEL 

- 


Bare Gro 
Urban 


C5 

B 
V 

o. 
O 


u 

1 


1 


on 

S 
E 

!/3 



372 



-o 

3 

_n 

c 
o 



1 



e „, 


^c 












portio 
ide thi 


3 


© 

On 
©n 

mwm\ 


o 
o 


On 
O 


o 


o 


c g 














- e 


rz« 












- * 


Gfl 












s v 


■ 












oportio 
tside tli 


o 
< 


© 
©N 

©n 


ON 

On 

ON 


On 
On 


o 

o 


o 
o 


u 3 


~ 












fr. 


Gfl 












B a, 


B 












oportio 
side thi 


o 

< 


K 

ON 

— 


O 

o 


© 




o 


u, s 


rz 












o. — 


35 












§■2 


■ 












roporti 
utside 


O 




NO 

On 


NO 


ed 


o 


«< 


On 


Os' 
On 


On 

ON 


~s 


© 


§ 












fi. O 


55 












s » 


s 












.2 j= 


— 












port 
idet 



< 


in 

©n 

— 


On 
O 


© 
© 


o 


o 


o 2 


s 












(5 ~ 


55 














# fi 












o — 














roporti 
utside 




< 


IT, 

ON 


Ov" 
On 


00 

ON 

On 
On 


o 

o 


o 

o 


• mm 












0. Q 


M 












■ 














■ 














« 
























1 




U 








O 


B 
3 


u 


B 








B 


-*— 


O 
"■C 

2 






.a 


0. 

B 


O 
E 

O a 


OS 


o> 






w 


C8 


CJ « 


B 


£ 






*5 

E 


"5. 




a. 


> 






a. 


S 


S3 - 


O 



373 



a 

03 !/i 

£ J2 

GO O 

D v£> 

<" ^« 

o 2 

* 1 

£ 

1 £ 

2 o 

S o- 
•S E 

On _ 

T< 

t*- 03 

•"* w 
•^ -o 
rt x: 

2 1 

— <U 

I i 

s ?? 

.2 U 

8.9 

E eef 

5 g 
u « 

O d 

g>« 

> c 

c <u 

II 

<*- E 
o •- 

o — 

■4-* •»- 

03 W3 

c2 c 

. 3 

~ "O. 

"" ' Xl 

! 3? 

H -5 



2 oo 

•Q.S 



8. 






03 

O O ") 
60 



x: 



C3 



2 «H 



S <t> Eo 



S 3 =§ 

b S 3 

r= > .E 



r- t; 






E 

E 03 ■£! 

« ^ 5 

.22 a ~ 

E I g> 

S 5 1 

C w. O 

C -n u 
O 1? O 

<** ,.5 

1 2 c« 

002 « 

c o 2J 

D. & bC 

c on -s; 
E - U 

2 E :s 
111 

111 

03 a> 

> o C 

111 

g oS 

.2 j= o 

_g x: °- 
"5 o « 

5 E-S 



-o 



E 
o 
u 
c3 

C 
03 

a, 

03 

E 

r^ 

r- 

— 



-o 

o 

u 

u. 

e- 

c 

+-* 
o 

c 

<L> 

i- 
U 



(U 

^-» 
03 

C 

O 

T3 

C 
03 

C 
03 
X) 



T3 

c 

I 

o 

u. 

QQ 



C/J 

C 

o 



« 03 

£ 2- 

u E 

c © 

- o 



o 






















* s « 






















9\ R DC 

— E e 






(N 


© 


00 


rn 




(N 


© 


00 


977- 
Ove 
Cha 






S 


■*' 


so 


vm 




r»' 


■<3- 


NO 






r- 


r- 


VO 


(N 




r~- 


r- 


NO 






















w* 












































■*■ 






















© OJD S 






















» « B 






















977-19 
Chang 
ccurrii 
ide the 


O 




ON 


o 

NO 


NO 
00 


© 




00 


© 


© 


-< 




On 


On 


r- 


ON 




m 


CN. 


-OS 






















mm 


























< 




















1977-19 
Chang 
Occurri 
Outside 




o 
o 


<N 


On 


fN 




On 


o 


00 


'& 




h» 


Os 


V> 




r- 


t^- 


NO 






















t» 






















t- — 4> 






















ON S M 

pM E s 






CO 


© 


(N 


r» 




00 


o 


<N 


952- 
Ove 
Cha 






<N 


NO 


f«i 


00 




«N 


NO 


m" 






<N 


<N 


(^) 


r~- 




fS 


(N 


m 






















i-H 


































































r- MJ^ 






















f*~ « 9 . , 






















ON ftX'E « 

7 g u 5 


— 






« 


O 


TT 


00 




in 


t^- 


— 


«S S 3 « 

88l| 




(N 


'^ : 


(N 


ON 




NO 


ON 
NO 


ON 
ON 


M O 9 






















■N 
























o 
< 




















1952-19 
Chang 
Occurri 
Outside 




© 


00 


«o 


00 




— 


© 


On 




© 


O0 


O 


•* 




H 


«r> 


w—< 


«5 




(^) 


■<r 


ON 




CM 


(N 


m 






















en 

■ 




5J 


*-» 










^- 






M 




E 


in 








~e 


I 






U 

e 
.o 

08 




.5 

f 

8 


E 
o 
b 

•a 

B 


w 


.Si 


B 
- 


IP 


E 
o 

B 


f> 


.a 


en 
> 




2 


Xi 
3 


*C 


e 

X? 






Xi 
3 


"u. 




Ci 


a> 


— 


"S 




<&. ^j 


4> 


E 


'S 




3 


£ 


Xi 


E 

- 




^ 


Xi 
CZ5 


a. 



374 



© 

on 5 m 
•- e 



ON 



ou 



On & .5 g fi 

rf. 3 s v 2 

-OS 



o 

On 
On 



ON 



o 



OX) 



o 



S 33 

11.1 
U 5 - 



oc 3 



o 



K as 4» 

ON 5 H 
^« E B 















oo 



00 



3 
C 

1 



I 



























t» 




ec 


x 






K 


V 


c 








ON 

1 

in 

On 


ox 

e 
a 

u 


i- 

s 
w 
o 




s 



< 


NO 






e 












— 








V 


6£ 




o 

< 

35 




On 

1 

ON 


ox 

B 
- 

U 


- 

3 
u 
w 

o 


s 

3 


00 






o 










w 












w 












« 












u 






V 






B 






e 






O 






£ 






i 






"O 
B 
OS 






$ 






"a. 






> 






- 



375 
occurred during 1952-1977, and at a slower rate during 1977-1990 (Table 3-11). The 
fires of 1954-1955 were probably responsible for some of the change from wet forest to 
shrub and prairie; these areas were rapidly re-vegetated with shrub-prairie and scrub 
associations by 1977 and wet forest-shrub types by 1990 (Table 3-9). Fires occurring in 
this area during the past century appear to temporarily arrest forward succession of 
vegetation, but in the absence of repeated, severe fire, recovery quickly occurs. The sill 
may actually accelerate drainage of this area during extremely high water levels by 
reversing the hydraulic head near the creek-river junction. This difference decreases as 
water levels drop (see Regional Hydrologic Trends). However, the sill does not stop fires 
in this area of the swamp. Fire activity in the southern third of the swamp has exceeded 
that in the remainder of the swamp during the "with-sill" period (see Chapter 5). 

Discussion 

Model Performance 

The Okefenokee Swamp hydrologic environment is well-represented by HYDRO- 
MODEL. Although some discrepancies among modeled and recorded data exist, the 
model scale is sufficiently detailed to permit assessment of the effects of the Suwannee 
River Sill on the Okefenokee Swamp landscape. In most cases localized failure of the 
model can be attributed to specific model or data features. The model processing area is 
the Okefenokee Swamp National Wildlife Refuge boundary and the Suwannee River 
floodplain to Echols County, GA; this limit was required by computing constraints and 



376 
project objectives. Features that may affect hydrologic processes in the swamp, such as 
watersheds of the northwestern streams, topographic relief beyond the eastern perimeter, 
and seepage flows along the eastern rim, were eliminated by this boundary. If the 
processing boundary had been extended beyond the refuge boundary, these features 
would have had an opportunity to influence swamp hydrology, which might have 
improved model performance in these areas. However, additional processing power, 
memory, and storage space would be required to model a larger area. 

Error can also be attributed to model processing scale, which was limited to 
500x500 m grid cells. Smaller scale would have improved local accuracies, but would 
have increased processing time and storage space exponentially. In most cases the model 
output agreed with interpolated check station data, which generalized local variability to 
the model scale. Scale errors occurred primarily where check station locations did not 
represent the general local environment. For example where recorder stations were 
located in ditches or holes, modeled water depth estimates were low because the grid- 
cells of the model topography did not represent the depression at the small scale. Areas 
of the swamp could be subset to create smaller regions for model processing, and these 
could be gridded at a smaller cell size. The smaller cell size would permit more local 
landscape detail. Re-sampling swamp topographic elevations at the smaller scale would 
be necessary, however, to provide data for a more resolute topographic surface for local 
models. 

Several assumptions were made in designing the model and creating the model 
data grids. As mentioned above, the swamp topographic surface was approximated with 



377 
data collected at a variable scale and interpolated to 250,000 m 2 grid cells. Local 
variability is lost at this scale, but hydrology at the landscape level is represented with 
sufficient accuracy. Other data used in the model were also interpolated from stations 
scattered throughout the area. Precipitation was represented by data interpolated from 22 
stations and totaled over semi-monthly intervals. As discussed in chapter 2, these data 
were at sufficient temporal and spatial density to accurately represent the area rainfall 
variability, so that model error is probably not attributable to precipitation estimates. In 
fact, 14 stations would have adequately represented area precipitation variability for 
semi-monthly estimates of daily precipitation (Chapter 2). Evapotranspiration rates were 
also estimated from interpolated point data that had been totaled monthly and then 
halved to approximate semi-monthly ET. These values were estimated using 
Thornthwaite's approximation of potential evapotranspiration, which has been reported 
to be a low estimate (M. Focazio, USGS unpublished data), although Yin and Brook 
(1992a) believed it was the best estimator of actual evapotranspiration in the Okefenokee 
Swamp when compared to Blaney-Criddle and Holdridge PET estimation methods. The 
model's ET adjustment of 25% approached the estimates of Thornthwaite's deficiency of 
30-40%. If model error is attributed to ET estimates, this probably is important primarily 
in the eastern swamp, where ET plays a greater role in directing water level fluctuations 
than in the western swamp. ET rates were also adjusted with a proportional multiplier to 
account for differences in local vegetation type. This modifier could be improved by 
species-specific ET rates instead of those for structural types (tree, shrub, prairie, open 
water); these values were not available when the model was constructed, but could be 



378 
incorporated in future model manipulations. Inflow and outflow volumes were also a 
source of error. They were estimated from measured flow rates and linear regression 
relationships between creeks and rivers. More accurate volumes could be estimated with 
a longer flow database that includes a greater range of water level conditions at more 
inflow and outflow locations. Additional recorder stations are also needed in the seepage 
flows along the eastern swamp perimeter to assess connectivity to swamp and upland 
water level fluctuations. 

Groundwater exchange was assumed to be minimal when the model was 
constructed (Hyatt 1984, Hyatt and Brook 1984, Blood 1981, Rykiel 1977), and the 
contribution of this component was considered less than the expected model error. 
Although this may be true from a landscape perspective, this assumption is probably 
erroneous on a local or basin scale. There is anecdotal evidence of subsurface 
contribution to the eastern swamp hydrologic environment; springs or upwellings have 
been reported near Floyd's Island (J. Burkart, pers. comm.) and in southern Chesser 
Prairie (pers. observ.), and elsewhere in the swamp (Hyatt 1984, Hopkins 1947). The 
minimal variability of water surface elevations in the Durdin Prairie area may be due to a 
subsurface contribution. Refining the volume and variability of this input should be a 
priority of future research on the swamp hydrologic environment, especially in the 
eastern swamp. 



379 
Effects of the Suwannee River Sill 

The Suwannee River has elevated water levels and prolonged flooding in 
approximately 18% of the Okefenokee Swamp. Most (15%) of the impounded area is in 
the Suwannee River floodplain to the east of the sill; 3% of the impacted area is in the 
Cypress Creek watershed where the effect may be due to backwater impounded rather 
than direct flooding by the sill. The actual linkage between the sill and this watershed is 
uncertain. This watershed is isolated from most of the swamp by landscape topographic 
features, such as sand-based islands (Blackjack, Honey, Billy's, Strange) and peninsulas 
(Pocket, Soldier's Camp) surrounding the Cypress Creek-Sapp Prairie area. Sweetwater 
Creek and Honey Prairie, located to the north and northeast of Cypress Creek and 
between the sill and Cypress Creek, have not demonstrated any change in water surface 
elevations or hydroperiod frequencies since sill construction. In addition, the impact 
exists under all water level conditions, unlike closer areas that are affected primarily at 
high water levels and less so at average and low water levels. There is a reversal of 
hydraulic head that occurs in this watershed under extreme high water level conditions 
that is less frequent and begins at higher water levels in the sill's absence. Removal of 
the sill creates a change in the river-creek profile. Cypress Creek levels drop below those 
at the river and drainage slows until the river recedes. 

The spatial extent of the sill's impact varies with general water level conditions. 
At extremely high water levels the area encompassed by Craven's Hammock to Floyd's 
and Sapling Prairies to the western edge of Chase Prairie to SCFSP and the Sill has 



380 
elevated average water levels of a few centimeters to a meter. At extremely high levels 
water in the eastern swamp (Buck Prairie to Chesser Prairie and south) that usually does 
not show surface movement may be forced to the east and southeast by the delayed 
drainage caused by the sill. The result is a head reversal maintained until the water 
impounded at the sill decreases. During high, average, and low levels, the eastern area 
shows little surface water movement and no difference in elevations with and without the 
sill. Water flow becomes apparent west of Coffee Bay as water drains out of 
southwestern Chase Prairie towards Billy's Island. These conditions probably occurred 
prior to the Suwannee Canal construction; attempts to drain the swamp towards the 
Suwannee River would probably not have been made had there not been evidence in the 
eastern swamp of westward flow, which begins in the region of the canal's westward 
terminus. 

At average and low water surface elevations the area affected by the sill's 
impoundment decreases. Under these conditions no differences in water surface 
elevations with sill removal occurred at Coffee Bay, Chase, Territory, Durdin, and 
Chesser Prairies, Double Lakes, Gannett Lake, Moonshine Ridge, and SCRA. This 
region is delineated in the surface topography by a rise that is not exceeded by the 
impounded water. Average water surface levels at Craven's Hammock to Floyd's and 
south Sapling Prairies to SCFSP decline with removal of the sill, although this drop is 
small in Floyd's and Sapling Prairies. At extremely low water surface elevations, the sill 
impounds water only in the riverbed and not in the surrounding floodplain, and it's 
removal has no effect outside the riverbed. 



381 
Even if all water flowing in the Suwannee River at the sill was retained in the 
swamp, the affected area would not exceed that affected at the highest water levels with 
outflow. The Craven's Hammock to Floyd's Prairie to western Chase Prairie to Billy's 
Island area would experience an increase in water surface elevations of 0.05-1.0 m, with 
greater increase closer to the sill. This affected area follows a topographic contour that 
would have to be exceeded by the impounded water surface elevation before a larger area 
of the swamp would be flooded. 

Greatest spatial impoundment by the sill occurs during high and extreme high 
water levels, which usually occur in the winter months when evapotranspiration is 
minimal and winter frontal systems bring precipitation. This is a period when wildfire 
frequency is low and lightning ignitions, which are the predominant cause of swamp 
wildfires, are least frequent (see Chapter 5). If the sill is to provide fire and drought 
protection, it should be increasing hydroperiods during drier periods when rainfall is 
minimal and lightning-caused storms are frequent. Because the swamp hydrologic 
system is so tightly liked with area rainfall and evapotranspiration, however, the sill can 
not impound enough water during the period when its impoundment effects are most 
needed to counteract drought and arrest wildfire spread. Even if all out flowing water 
were captured by the sill, the region impounded under low water conditions would 
increase minimally. Increasing the affected area to significantly more than 15% of the 
swamp can only be achieved by increasing area rainfall. This effect occurred during 
February-March 1998, when record amounts of precipitation fell in the swamp watershed 
and throughout the swamp, and water levels reached record high levels. 




382 
The communities affected by this impoundment are undergoing the same 

successional changes as those beyond the reach of sill's direct effects, although rates of 
change were initially faster in the impounded region. Communities in the affected region 
were in general beyond the stage most effected by increased hydroperiods (germination) 
when flooded by the impounded water, and are undergoing conversion to shrub-forest 
and wet forest types. Removal of fire from these areas and the surrounding swamp has 
facilitated this change, which probably will not be reversed until a widespread, severe 
and extensive drought is accompanied by a severe fire. Under these conditions the sill 
will not stop fires from igniting and spreading, and the impounded area will be 
negligible. 
System Sensitivities 

Although the Okefenokee Swamp can be described as a "bowl in the landscape", 
there is diversity in the landscape that creates a spatially variable hydrologic environment 
that does not react uniformly to system perturbations. In this study five basins were 
delineated where hydrologic fluctuations follow general trends in seasonal weather 
patterns but levels of variability differ (Figures 2-1 1 and 2-12). Sensitivities to 
manipulations of components in the water budget and landscape vary with the basin. The 
eastern swamp basins quickly respond to changes in evapotranspiration rates, whereas 
the western basin shows less sensitivity to this parameter. More disruption to the western 
hydrologic environment occurs when semi-monthly outflow or inflow proportions are 
adjusted. These responses reflect features in the swamp landscape. The western swamp 



383 
is in the Suwannee River floodplain, and receives more throughput, with inflow from 
northwestern creeks and outflow in the Suwannee River. Evapotranspiration rate 
adjustments affect the volume of water entering this area in peripheral regions, but most 
of the area's water movements are due to river and creek fluctuations, and ET variability 
is of secondary importance, primarily as it influences the volume of in flowing and out 
flowing water. Fluctuations in river and creek volumes due to influences outside of the 
Refuge boundaries have the potential to disrupt the hydrologic environment of this area 
of the swamp. The eastern swamp contrasts the western swamp's low sensitivity to 
fluctuations in ET. Throughput is minimal under all but very high water level conditions 
in the precipitation-evapotranspiration driven eastern swamp. Small changes in ET in 
the eastern swamp cause large changes in water levels. In reality these changes could be 
caused by altered standing water volumes due to draining or flooding, or changes in 
vegetation coverage. It is possible that the small fluctuations in annual water levels of 
the northeastern basin are directed by species-specific ET rates; a change in vegetation 
coverage due to small alterations in water levels could result in species conversions in 
this area of high diversity as species-specific ET rates also change. 

The Okefenokee Swamp has a history of hydrologic manipulation. Although 
currently the most visible disruption, the Suwannee River sill has had limited impacts on 
swamp hydrology and vegetation compared to other manipulations of the past 100 years. 
The Suwannee Canal has probably affected flow to the western swamp in the adjacent 
areas, and may increase de-watering of adjacent prairies under certain conditions, and the 
canal berms are an interruption in what once was an extensive prairie (Christie to Grand 



384 
Prairie). These areas east of Bugaboo Island remain largely hydrologically isolated from 
the western swamp, however, in spite of the canal. Dredging in the northeast has also 
affected local hydrologic dynamics, but the extent is limited, which has isolated the 
impacts to the Kingfisher Landing-North Durdin Prairie region. Although the swamp 
seems to be recovering structurally from early 20th century logging, species composition 
may be changing to associations less dependent on fire for maintenance (see Chapter 4). 
Ultimately the hydrologic environment may change, as ET rates vary with different 
species, and different vegetation alters surface water flow rates. Perpetuation of fire in 
the landscape may also change with the predominance of species that do not readily carry 
fire. These changes are exacerbated by an altered natural wildfire regime, that is 
manipulated by fire management protocols to limit the severity and extent of fires, 
particularly in the swamp perimeter where many wildfires are naturally ignited. The 
intended function of the sill was to assist in these fire control efforts; it does not appear 
to be achieving this purpose throughout the swamp. However, in combination with the 
historic perturbations that have variously affected the swamp hydrology and vegetation, 
the sill is contributing to changes in driving processes that ultimately structure the 
Okefenokee Swamp landscape. 



CHAPTER 4 
LANDSCAPE LEVEL VEGETATION CHANGES IN OKEFENOKEE SWAMP 



Introduction 

Dynamics of Swamp Vegetation 

Plant community composition and distribution are the result of dynamic 
interactions of autogenic (self-imposed) and allogenic (environmentally imposed) 
factors. The sequence of species occupying a site is determined by the abiotic 
environmental conditions, interactions among individuals and species (such as 
competition for resources), and propagule availability. The result on local and landscape 
scales is not static; as species composition, structure, and site environment change, so do 
the structure and composition of the landscape, creating a "moving mosaic" of 
communities in various stages of development, responding to disturbances with which 
they evolved. Responses to these events may be fairly predictable; this predictability in 
response to change is due to the regularity of the types and intensities of disturbances or 
driving functions occurring throughout the system's history. Disruption of the 
relationships of community components and processes may alter the responses of 



385 



386 
individuals, species, and communities. A change in the composition, structure, and 
dynamics of the landscape may ultimately result. 

Vegetation communities generally undergo a predictable sequence of change in 
Okefenokee Swamp, determined primarily by site hydrology and fire history (Deuver 
1984a, 1984b, 1983, 1982, 1979, Hamilton 1984, 1982,Cypert 1973, 1972, 1961). 
Longest hydroperiods and deepest water levels are tolerated by species in areas of open 
water and aquatic prairie (e.g., spatterdock, Nuphar luteum; fragrant water lily, 
Nymphaea odorata; golden club, Orontium aquaticum). Shallower water depths and 
more frequent exposure characterize herbaceous prairie (e.g., Walter's sedge, Car ex 
walteriana; yellow-eyed grass, Xyris spp.; broomsedge, Andropogon virginicum; redroot, 
Lacnanthes carolimana). As water depths decrease and exposure times increase due to 
litter accumulation or drought, shrubs invade, with shade intolerant species (e.g., 
fetterbush, Leucothoe racemosa; titi, Cyrilla racemiflord) gradually replaced by those 
more suited to shaded conditions (e.g., hurrahbush, Lyonia lucida). Forest species 
tolerant of longer hydroperiods and deeper water are cypress and blackgum, which may 
eventually be displaced by bays as ground surface rises due to litter accumulation, and 
shade creates less favorable conditions for cypress and blackgum regeneration. 
Depending on the burn intensity, fire can disrupt this progression and reset the 
community to an earlier stage, or retard the cycle by pruning above-ground growth, 
without changing dominant species composition (Hamilton 1984, 1982) (Figure 4-1). 
Evidence of these cycles exist in the peat throughout the swamp where community 



387 



4,5 



Prairie 



Prairie 

with 

Cypress 



Prairie with 

Cypress-Shrub 

Swamp 




1 



Cypress-Shrub 
Swamp 

| 3,5 

•Cypress-Bay-Blackgum 
/ Swamp 

1,2 




Mixed 
Swamp 




Sprout Growth 
Blackgum 



Prairie with 
Shrub Swamp 



Prairie with 

Bay-Shrub 

Swamp 



! 



2,3 



Bay-Shrub 
Swamp 



W 



Bay-Swamp 



/ bWi 

1,2 



1 = Light Fire 

2 = Moderate Fire 

3 = Severe Fire 

4 = Very Severe or 

Frequent Fires 

5 = Logging Followed 

by Fire 



Figure 4-1 . General effects of fire and logging disturbances on Okefenokee Swamp 
vegetation types, adapted from Hamilton (1982). 



388 
succession checked by fire has resulted in a perpetually changing landscape since the 
swamp's most recent peat accumulation began 6,500 years ago (Cohen 1975, 1973b, 
Cohen et al. 1984). 

Most of the Okefenokee Swamp has been affected during the past century by 
some type of man-induced manipulation, including logging (1890-1942), ditching ( 1 890- 
1900), peat mining (1930s-early 1950s), alteration of fire regime (1937-present), canoe 
trail maintenance including trimming and dredging (1937-present), and impoundment 
(1960-present) (Trowell 1989c). The spatial and temporal effects and permanence of 
these modifications are uncertain. Not only is community composition directly altered 
by these processes, but the subsequent responses of the landscape to these disturbances is 
also affected by the change in species composition. Predictions of responses to future 
disturbances, either "natural" such as wildfire and drought, or "unnatural" such as 
impoundment, draining, or modified fire regime, are less certain following these artificial 
disturbances. Hamilton (1984, 1982) documented Okefenokee Swamp vegetation 
composition in 1977 following the turn of the century logging and Suwannee Canal 
construction, peat mining of the 1940s, and Suwannee River Sill construction in 1960. 
He proposed sequences of changes observed due to these factors (Figure 4-1), and 
expected succession in the absence of these disturbances and with various hydropatterns 
(Figure 4-2). He did not propose a time line for these changes, since the frequency of 
these disturbances and the system's response had not been examined over a sufficient 
period. The availability of GIS permits a spatial comparison of the disturbances and the 



389 



Prairie 




Prairie with 

Cypress-Shrub 

Swamp 



Cypress-Shrub 
Swamp 



Cypress-Bay -Blackgum 
Swamp 



O 




Prairie with 
Shrub Swamp 



Prairie with 

Bay-Shrub 

Swamp 



Bay-Shrub 
Swamp 



Bay Swamp 



Mixed Swamp 



VJ 



Succession Drivers 


• 
• 

o 


= High Stem Density 
= Deep Water 
= Shallow Water 
= Short Hydroperiod 
= More Shade 

= Less Shade 



Figure 4-2. Autogenic succession and conditions that drive succession in the Okefenokee 
Swamp, adapted from Hamilton (1982). 



390 
system's response during the century since the initial logging occurred. The recent 
periodicity of the resulting changes might be elucidated with this type of spatial and 
temporal analysis. 
Sill Affected Vegetation Change 

In addition to questions of the Suwannee River Sill's effects on the Okefenokee 
Swamp hydrology addressed in Chapter 3 are questions of its effects on the swamp 
vegetation composition and distribution. Changes in swamp vegetation since the sill was 
constructed can not be attributed directly to the sill without also examining vegetation 
responses to these other disturbances, within and outside of the area hydrologically 
affected by the sill. This chapter examines swamp vegetation landscape composition and 
community distributions in the Okefenokee Swamp prior to logging (1850-1890), prior to 
the wildfires of 1954-1955 and sill construction (1952), and 17 (1977) and 30 (1990) 
years following sill construction. These intervals were selected due to data availability , 
as well as timing of fire, logging, and sill construction. My objectives were to identify 
the types and locations of vegetation change occurring in the swamp during these 
intervals, determine the roles of various disturbance types (such as fire, logging, and 
impoundment) in directing the changes occurring in the swamp landscape, and assign a 
temporal scale to these changes. 



391 
Methods 

Logging Tramlines 

Most of the marketable timber (including slash pine, Pinus elliottii; longleaf 
pine, P. palustris; pond pine, P. serotina; pond cypress, Taxodium ascendens; sweet bay, 
Magnolia virginiana; swamp blackgum, Nyssa sylvatica v. biflora; loblolly bay, 
Gordonia lasianthus) was removed from Okefenokee Swamp during 1890-1942 along 
railways (tramlines) constructed to transport timber from the swamp interior; where the 
peat surface was inundated, the rails were elevated on pilings above the water surface 
(Trowell 1994, 1984b, 1983, Izlar 1984). Most harvesting occurred within 100-300 m of 
the tramlines (Trowell 1994, 1984b). Although the railways were dismantled following 
timber removal, evidence of the harvest remains, including landscape-level scarring in 
the vegetation structure and bases of pilings exposed during low water periods. Trowell 
(1994, 1983) compiled a map of tramline locations throughout the swamp using logging 
company records, aerial photographs, and survey notes. This logging tramline map was 
not georeferenced to a coordinate system; in order to compare vegetation, fire, and 
hydrologic history between logged and unlogged areas, the tramline map needed to be 
referenced to a coordinate system common among the maps. To create the geo- 
referenced tramline map, historic logging rails were extracted with ARCINFO (version 
7.0, ESRI, Inc., Redlands, CA 92373) from USGS 1:100,000 Digital Line Graph (DLG) 
coverages from 1994, and the resulting coverage was checked against 1994 USGS 



392 
1 :24,000 quadrangle maps for missing and mislabeled logging rails. Missing rails were 
digitized from paper USGS 1 :24,000 maps and merged with the logging rails derived 
from the DLG coverage. Trowell's 1994 map, "Logging Railroads in the Okefenokee 
Swamp (1889-1942)", was photo reduced to 8.5x1 1 inches and scanned at 300 dots per 
inch on a flatbed scanner. The reduced, scanned map was georeferenced to the tramlines 
compiled from the USGS map. An affine transformation was used with 8 control points 
to match the maps. The affine transformation function is 

*'=Ax+By+C 

j'=Dx+Ey+F 
where x and y are coordinates of the input coverage (the original tramline map) and*' 
andj' are coordinates of the output coverage (the USGS map). A, B, C, D, E, and F are 
computed by comparing the differences in positions and locations of the control points 
between the maps. The control points are scaled, translated, and rotated between maps 
to achieve the match (ESRI 1992), and the functions are then applied throughout the map 
to complete the transformation. A root mean square error (RMS) of 150 m was declared 
acceptable; this was at least as accurate as the original tramline map. Control points 
were reselected until this accuracy was achieved. Artifacts such as text scanned from the 
original map were removed before the scanned, raster version was vectorized. Breaks in 
logging lines, errors introduced in the transformation process, or missing lines were 
corrected using the following decision rule sequence: 

1) The USGS logging tramline information was believed to be planimetrically 
accurate, but less quantitatively accurate than Trowell's original tramline map. 



393 
Therefore, if the original logging tramline map indicated a railway, that feature was 
included in the digital map. If the feature was mapped by USGS, the USGS feature was 
retained and the original map feature was removed, since in the transformation its 
position was less accurate than the USGS feature position. 

2) If a tramline was present on the original tramline map and not on the USGS 
map, the tramline information was checked against historic photographs (1952) and 
satellite imagery (1990) to locate existing large-scale marks in the vegetation, so the 
location could be placed on the digital map. The missing tramlines were digitized using 
scars (large-scale marks in the vegetation where recovery from logging was occurring) 
visible from these sources as guides. 

3) Where features existed on both USGS and original tramline maps, but 
locational discrepancies occurred, the logging tramlines derived from the original map 
were adjusted to the position indicated in the USGS map. 

Tramline locations in the final composite map were estimated to be within 400 m 
of their true location. This map became the base map, composited with island, stream, 
and river vectors, for compiling the pre-logging vegetation map of the swamp. 

The final tramline vector map (or arc coverage) was converted to several grids 
representing logged and unlogged areas. A buffer was used around the logging tramline 
arcs to represent the logged and unlogged areas separately (Figure 4-3). All areas within 
200 m of the logging tramlines were labeled as "logged", and areas beyond this 200 m 



394 




GO 

_o 

3 

E 

M 

s 

a 
a 

■ 

o 
•** 

•o 
<L> 

VI 

3 

£ 

3 
X> 

T3 






«J 



en 

O 



C3 

u. 

50 
5 

'Si 
oo 

c> 

•— 
o 

V3 

c 
o 

1 

u 

o 



m 
i 

V 
■*- 
3 
&0 

E 



395 
buffer were considered "unlogged". The buffered tramlines were gridded to 10 m, 240 
m, and 320 m grid cells using ARCGRID, for comparison with other vegetation maps 
(see below). 
Pre-Logging Vegetation ( 1 850-1 890) 

The map of logging railways created from the tramline features on the USGS 
1994 DLGs and Trowell (1994, 1983) was used as a base map to plot vegetation 
community distributions along the pre-logging survey routes. Although much of the 
Okefenokee Swamp was logged during 1890-1942, and the composition of the harvest 
was estimated by Hopkins (1947) and Izlar (1984), a spatial representation or map of pre- 
logging vegetation map had never been constructed. An approximation of pre-logging 
vegetation composition occurring during 1850-1890 was made from notes compiled by 
Trowell (1994, 1989a, 1989b, 1988a, 1988b, 1984b) of several surveys conducted during 
1 850-1942 (Table 4-1 ). Because no detailed maps of vegetation cover were supplied 
with the summaries of the surveyors' narratives, positions of vegetation types were 
approximated from the survey descriptions of location, distance covered since last known 
position, and notes compiled in survey route descriptions by Trowell (1989b). 
References to landmarks such as streams and rivers, large islands, and large prairies also 
provided positional information (Figure 4-4). Although there was room for error in this 
method, there were consistencies among the surveyors' descriptions of areas visited by 
more than one surveyor, and distance estimates were generally comparable among 
surveyors. It was assumed that the vegetation currently most likely to be different from 



396 

Table 4-1. Sources of pre-logging survey notes used to create the pre-logging vegetation 
map of Okefenokee Swamp. 



Survey Party 


Survey Date 


Reference 


Mansfield Torrence 


1850 


Trowell (1989) 


Pendleton-Haines 


1875 


Trowell (1989) 


Constitution (Clarke, 
Pendleton, Haines, Little) 


1875 


Trowell (1989) 


Fremont 


1878-1879 


Trowell (1989) 


Roland Harper 


1902, 1919 


Trowell (1988) 


Suwannee Canal Company, 
Hebard Lumber Company 


1890-1937 


Trowell (1984) 


Various Logging Interests 


1895-1942 


Trowell (1994) 









397 



1 
I 

s 

c/) q: #q_ w a. u: 



11 
If 

0)1, 

|-t— r-r\ 






</) 

0) 

■s 




398 
historical, pre-logging descriptions was where the tramlines, and hence logging, had 
occurred. However, survey information included areas later logged as well as those 
never logged, so the final pre-logging map includes all areas described by surveyors. 
Dates of logging were also noted from Trowell (1994), or references in the survey notes 
(Trowell 1989a, 1989b, 1988a, 1988b, 1984b) and logging company records (Trowell 
1984b). After vegetation descriptions were recorded on the tramline map, they were 
summarized into 12 classes (Table 4-2). The selection of vegetation class types was 
determined by the 1990 satellite image classification (see satellite image classification 
discussion); a common set of classes among vegetation maps was necessary to permit 
comparisons among maps for change assessment. Boundaries of vegetation communities 
were estimated on a paper tramline map and screen-digitizing on the tramline coverage 
to create vegetation polygons (Figure 4-4). The polygons were converted to 10 m grid 
cells in ARCGRJD, and compared with the other vegetation maps in IMAGINE (version 
8.2, ERDAS, Inc., Atlanta, GA 30329) using the MATRIX and SUMMARY procedures 
(ERDAS 1995). 
Post-Logging Vegetation (1952) 

Estimation of vegetation community distributions prior to sill construction and 
the extensive wildfires of 1954-1955 was made from SCS 1:24,000 black and white 
aerial photograph stereo pairs taken during March 1952. Most of the swamp was 
included in the flight lines of this photograph set. The area north of UTM-Y= 3425000 
was not recorded during March 1952; this region was included in flight lines flown 



399 



o & 



U 
































o *. 
































<a <s 




c 




C 
























SP« ft 




ea 




CO 
























WD cb 




Xi 




X 
























3 "O 




c 




c 


en 






en 




en 


en 




en 




en 




■ 

-a 




T3 








1) 
E 




CD 

E 






i 




CJ 
E 


o v 




c 




C 


o 






O 




O 


o 




o 




O 


s- B 




3 




3 


u» 






ft 




ft 


ft 




u. 




ft 


« s 




O 




O 

E 


•o 






T3 




T3 


"O 




T3 




■g 


5 £ 




a 




a 


9 

a 






C 
J3 




a 


a 




C 
^co 




c 
^co 


C6 c 

6 U 




<u 




u 


■*— » 






■«— ' 




— < 


"*—• 




•*-» 




■*-• 




e 




c 


<L> 






<u 




<D 


o 




CJ 




CJ 




CO 

CO 




CO 

CQ 


£ 






£ 




£ 


^ 




^ 




^ 


.o ■ 






























X 


««- «N 






























5 

X 

GO 

i 
en 


upings 
d Map 




c 

a 




c 

cct 
























£ 




•e 






















en 
CJ 


ssGro 
mbine 




T3 




-6 


0) 

E 






+-» 

en 
CJ 

E 




i 


2 




en 

cu 

s— 




E 

9 




C 




c 


o 






O 




o 


o 




o 






3 




3 


ft 






ft 




ft 


fc 




u- 




CO 


5 U 




s 




8 


T) 






T3 




•o 


T3 




"O 






o 




o 


1 






J 




a 


C 

_C0 




c 

_e0 




CQ 






s 




s 


+-* 






*-* 




*-» 


+2 




*3 




E 

3 

o 






CQ 




CO 

CQ 


<D 






D 




u 


CJ 




CJ 










£ 






£ 




£ 


^ 




^ 














• r\ 




















>i ^ 










c+i 


b 






CD 




en 
en 




•«-» 






CO -g 










o 


o 






c 


en 


C/5 


c 




X 


■ 

o 

ft 

E 

u 

■ 




c 

CO 

X 




i 

•c 

'O 

CO 
P 


i 

fca 

CJ 
> 

O 

C 


60 
en 

<L) 
U 


VI 

X 


'ft 

<u 

E 

CJ 

1 


X 

2 

X 
en 

£ 

3 


i 

o 

C 
CO 


| 

X 

4— » 

c 
co 

c 


CO 

c 

1 

o 

"O 
X 

3 


i 

C 

"ft 

CJ 
00 
T3 


1 

CO 
X 

E 
| 

O 

CO 


-*= E 

1 1 

? o 

•E ■* 


a 

■ 
■ 




3 

•a 

c 


c 


en 


'ft 

T3 

C 

O 


6 

X 


t/2 


en 

co 


00 

o 

CO 


E 


1 

o 


en 
en 

X 

s 

X 
en 

CU 
O. 

CO 

E 


CJ 

-a 

a 


C en 

1 i 


es 

U 




3 

1 

I 

CD 


E 

a 
o 

CJ 

> 

0) 
T3 


ts 

T3 

CJ 

jg CO 

Oh £ 


c 
o 

X 
en 

a 

09 


3 
00 

o 

CO 

X 


1 
o 

1 

Vi 


X 
X> 

"o 

— 

X 

o 


X 

en" 
en 
<D 
hi 
Cl 
>% 
U 


o 

6 

X 
o 
C 

<u 

00 

O 


E 
1 
o 

CO 

s 


'a 
2 


X 
i 

en 
en 

i 

CJ 


P en 

i if 

co ft 
CQ & 


ft 
































es 
































■ 
es 


& 
























en 

T3 
O 

o 




X 

2 

X 


en 
en 


1 


















en 






V 




i 

en 


es 


R 


•e 




8 


<u 










en 
U 


en 

>> 




c 

CO 




en 

eU 


U 
c 


-5 

es 






co 

-J 


a 
ft 






>> 




E 


CO 

CQ 




1 




E 
ft 


B 

es 

6 

em 

o 

> 


I 
1 

ex 
ex 


c 

3 

o 

1- 

O 

eO 




a 

E 
3 
■*-» 

3 

1 


■♦-» 
6 






co 

CQ 

x> 
*o 

X 

o 




u 

1 

u 

X 
o 

CJ 

CJ 
00 


CJ 

"S. 
CO 

1 

E 

3 




1 

ft 

o 

1 

CJ 

c 




u 

1 

CQ 
E 

3 




•^ 


CQ 




< 






►J 




O 


o 




ft 




o 



400 



1~ 

e 


£ 


«*- 


<S 


■ 


a. 


ejj 


95 


'EL 


i 


3 
O 


1 


U. 


B 


o 


£t 




E 



5 U 



<D 

i- 

o 

b 

T5 

c 
•is 

1) 



C3 

c 

I 

O 



0-, 



I 



•g 



00 



X> 
OO 



X) 

3 
u. 

X! 

OO 



X 
00 



u 


u 


B 


c 


■^* 


•*M 


a. 


Oh 


T3 


T3 


C 

C3 


S- 






a. 


a 


D 


D 



k 

C ■ 
CS PN 

g) s 

5 « 

s- c 

OS 

S i 

95 C 

5 U 







ea 




£ 


<u 


c 


•c 


1 


2 


O 


Qh 



.2 
'C 

i 



'C 



X 

00 



X 

2 

X 

OO 



X 

3 

J 

00 



oo 

03 

CQ 



s 



c 

o. 

D 



c 



T3 

C 
93 

"S. 

D 



u 
(J 
en 
0) 

o 

St. 

IT: 

9 



X . 




.- b 




w O 




ypre 
derst 




<D 3 


1 


atur 
rub 


C 
1 


A 5/3 


O 



<u 



ra 



;g 



S ex 
o 

'■^ 
73 

3 
or 
cs 

Q. 
6 

u 

Q 



3 
O 
<D 

u 

93 
| 



o 

o 
•o 
c 

O 

8 

93 

SL 

00 



<u 

93 



■5* x 

C 93 

S J 

3 93 

E 3 
11 

oo y5 



2 
'C 

x 

c 

X -C 
<- CO 



i e 
1 1 

1 s 

X O 

00 O 



c 

93 

"o. 

3 
i 

T3 

C 

a 

o « 

3 <- 
C U 

oo .5 



X) 

2 

x 

CO 

X 



^| 

o 2 
•5 u 

X T3 

9 a 

J 3 



X 

s 2 

, CO 

Cm K 

C S. 

O 60 

» 2 

o s 

Z u 

i« E 

a is 

oo S. 



c 
o 



3 
O 



5 bo 

o -S 

iJ 93 

U n. 



o 

3 

_n 
"■^ 
c 
o 
o 

■ 

■ 

i 

<U 



D. 
SB 



a 
ec 

ee 
en 

95 






V2 



1 



00 




^ 


CO 


E 

3 




I 

C/3 


CO 

co 




E 


93 
E 


Q 
i 

VI 


€ 
^ 




o 


1 


c 


60 


■♦-» 

93 




a. 


-a 


3 

5" 


U 


O 


00 


< 







X 


Jz> 




2 


hJ 




-C 
00 


13 


X 
00 


1 
b 

.93 

'C 

CQ 



00 

1 

I 

i 

D 
•a 

Q 

X 



c 



» 



00 

CD 



c 
cC 

•a 

c 

a 

c. 
D 



00 
I 

S 

u 

C3 

u 
U 



401 



c * 






«S <s 






& cu 






S s 












3 X 
o a> 

- s 


c 


o 

c 


O 2 


Cu 


Cm 


8 ? 


13 

C 


x 

c 


« r? 


CO 


CO 


u° 






5. 


Q. 


D 


D 





CO 


H 


CO 


en 


V) 


CO 




CD 


u 


B 


CJ 


u 


CD 


o 


c 


e 


E 


c 


E 


W 


c 


o 


o 


o 


o 


O 


O 


£ 


Uh 


U- 


tu 


MM 


HH 


Um 


X 


X 


XI 


X 


X 


X 


X 


C 


c 


c 


fi 


C 


C 


C 


cO 


03 


a 


cO 


co 


cO 


CO 




SI 


••H 


■*—» 


-*-* 


■♦n 


o. 


(L> 


u 


b 


6 


C 


CD 


£ 


£ 


£ 


£ 


£ 


£ 



© " 

M- 1— ■ 



CO 

DC 



C 

a* 

3 X 
O V 
U B 

OS 

* E 

6 U 



c 

£ 

X 

c 

CO 

"5. 
D 



c 

£ 

X 

J 
o. 

D 





CO 




CO 




CO 


co 




CO 


co 






a) 




<D 




B 


SJ 




CD 


B 




l) 


c 




l_ 




E 


E 




E 


E 




c 


o 




O 




O 


O 




O 


O 




£ 


u_ 




mm 




mm 


Uh 




mm 


HH 




x 




XI 




X 


X 




X 


X 




x 


c 




c 




c 


c 




c 


c 




c 

03 


CO 




03 




cd 


CO 




cO 


CO 




-*n 




-*-» 




•*H 


•+-» 




-*-> 


•*H 




D 


cj 




1> 




CD 


CD 




6 


u 




s 




£ 




£ 


£ 




^ 


^ 




1 

CD 


1 




c^ 




CO 

CD 


X 

u 

l_ 
CD 

S 

CO 
u 

CO 




co 


♦l 




X 

DO 

©" 


x 

CO 

X 
fi 

co 




X) 

£ 
3 
DO 




CD 
E 

DO 

C 
3 




CO 

U 

E 

O 


Bo 
J 

X 




cj 
E 

cO 

B. 


X 

cu 

c 


b 

o 

■*h 

CO 

u. 
<U 

X 

C 
3 


-X 
O 

co 

X 
co" 

C/3 
<D 

E 




O 
>> 

x 
■*-» 

X 

CJ 

X 


X 

CO 

>-. 
CO 

Xi 


CO 

1 

x 

CO 

X 


X 
cu 
E 

CJ 

6 

CO 

u 

CO 

X 


co" 

CO 

X 

X 

■*H 


co 

X 


a. 
x 

CO 


a 

X 
co 

So 


>> 
CD 

ej" 

c 


1) 

u. 
3 

•4-4 

X 

1 


i 

c 

Cu 


5 
D 


c 

CO 
X> 

"o 

X! 


co 

>> 

co 

03 


CO 

co 

CD 
E 

EL 

U 


X 
co 

<s 

c 
'5. 



J. 

CO 
0) 

Q 

CO 
CO 

CB 



1 ^ 

a. o 

•a i 

CO u. 

CD CD 

"5b g 

c B 

o 3 



I'll 

C co 

if | 

CO CD O 

"co "5 <*- 

US U 

1 J= rS 

CD -tJ co 

Q ^ •> 



CO 

c*- X 
CO 3 
_U C 

DO •£ 

J? 

H. JO 

% 

X 3 
co -> 

j i 

co C 

g .9 

S J 
5, 



VI 



x 

CD 

3 

i 

O 
o 

<N 
i 

CD 

I 



a 

CB 



X 

a 

CB 
to 

CO 

as 



CD 

> 



CD 

C 

CU 


CD 

C 


u 

CO 

c 

CD 


CD 

co 

Ih 

CO 

Q. 


a 


CO 



I 

I 








c 






CD 






CD 






i 






t- 






G 






> 






CD 






X 




CD 

C 

£ 


CD 
> 
CO 


CD 

C 

cu 


X 

c 


CD 


X 


CO 


X 


CJ 




CJ 


X 


o. 


CJ 




h0 


z 


>i 



CJ 

c 
Cu 
Xi 

2 

CD 

c/2 



c^ 

m 



CO 

co 

I s 

rj X 
i u 

>, x 



402 



3 
C 

O 

I 
i 

o 

I 



u 


























o - 


























•6 <S 


























SP« Ck 


























» es 


























s 2 

3 "O 


*-< 
co 


CO 


■*■» 

CO 




















2 


y 


>- 




















© V 


o 


o 


O 




















1- c 

OS 


to 


fe 


to 




















-a 


"3 


T3 




















Class 
Com 


c 


c 


C 




















_« 


J2 


■2 




X 


X 




X 


X 


X 


X 


X 


6 


6 


"*3 




00 


00 




S 

X 

oo 


5 

X 
00 


00 


oo 


2 

X 
00 


u 
o ■ 


























an Q. 
























u 


P ■ 
























■s 

'3 

t- 


























o $ 


•4-i 


*-* 


+■» 


















to 

1 


Class Gn 
Combin 


CO 


n 


CO 
















CO 


jb 


U 

E 
O 

C 


E 
O 

tin 

-o 

c 


o 

C 














1 


CO 

g 

U 


2 

X 

oo 
■ 

CO 

co 


a 


a 


.5 




x 


X 




X 


X 


X 


X 


y 




*3 




*3 




00 


oo 




oo 


I 

X 

oo 


2 

X 
00 


2 

X 

oo 


to 




00 


o 

■4-» 












p 




1 

Q 




co 
I 

I 
co 




o 


E 












8 


a 


JS 


8 


B 


>> 


6 

T3 








X 




o 


3 


O 


CO 

9 


9 

8 


£ 
o 


X 


x 

1 

CO 

X) 

— * 


c 

-o 
6 

E 

i 

o 

CO 

af 

o 

co «• 


CO 


'1 

s 


CO 
I 

| 

CO 

T3 

C 
J3 

i 


CO 

X 

"i 

CO 


g 

a 

•22 


co 

X 

CO 


co 

X 

'1 

CO 


C 

E 

5 co 

S x 


Q 

a 
8 


x 

-a 

c 


CA 

x 

CO 

co 


X) 


£ 

CO 


X 

s 

X 


'3 

E 

s. 


X 

2 

X 


X 

2 

X 


CO « 

X Cu 

~ E 
co "3 
co w 


es 

c 


CO 

CO 


£ 
1 


CO 

c 


2 

s 

3 
O 
>- 


co 

c 


13 

s 

es 


co 

-a 

c 


CO 

iu a. 




U 

to 4> 

u £ 


B 

i— 


.9 

o 
«a 

5 


.& to 

§ E 




J CO 

s u 

<L> c 

Bfc 'a. 


X 

2 

X 

oo 


C3 

■^ CO 

1) >. 


U 3 


D. 


























es 


























s 


























■o 


























B 

es 
























■c 


■ 
























•^n 


en 
























9 


es 
























B 


























to 


U 






















CO 


X 


B 
es 






E 






X 




u 

c 


| 


03 


co 
U 

E 


00 


6 




CO 
CO 


3) 






oo 




to 

1 


to 

I 


CQ 


u 

1 


CO 

co 


M 


X 


E 
O. 
>> 


^ 




X) 


X 




X 


X 


X 


X 


J> 


> 


S 






J 


2 

o 




I 


2 

X 


2 

X 


2 

X 






00 


U 


OQ 




00 


OO 




00 


oo 


00 


00 






403 



o * 
18 «s 

to a, 

S 08 

•s| 

a * 
a -o 

- e 

as 

08 ° 



CO 

u. 

o 

T3 

C 

a 

(L) 



u 

c 

5 

CU 



0> 

I 



c 



T3 

C 

a. 



co 

D 

o 

J 



X) 








H 


03 


x: 


t- 


OO 


Cu 



CO C. 

S> i 

a. m 

3 TS 

e v 

s- a 

OS 

*> fi 
03 ® 



U 



03 



O 

OO 



U 

cu 



u 

•c 

£ 

cu 



I 

I 



C 

£ 

T3 

C 
J2 

c 
D 



CO 

i 

O 
Pu 

"O 

c 

■ 



3 

x: 

02 



"E 

'5 

Cm 



co 
co 
03 

u 



E 

3 
60 

« .2 

s s 

"O D. 

S-s 

a. i> 

° 5 W 

60 O. <u 

I h -g 
o S 3 

>- 1 



03 
Q. 



V} <— 

<D 03 

to % 

•o ■ 

« c 

r- CO 

•g (U 

£ 60 

.8 1 

'3 £ 



o 

— B 

£ (S 3 



1 p 

03 C 



.2 ^T 

C o 

& § 

Si 
| & 

D. co" 

Q 3 



CO 

B 

8. 

T3 
B 
03 

CO 

cu 
C3 



CU 

I 

B 
| 

o 



T3 



3 
_fi 

B 
O 

o 

I 
I 

i 

u 

1 



s. 

08 



•g 
a 

08 

OB 
80 

05 



B 

2 

CU 



•S £ 

03 — 

J "5 

&0 60 

S 



P O 



CO 

00 P. 3 



B "~> 



o 

~ CO 
£j 

£l 

§> 2 

"o -S 
s w 

«. '£ CO 
CO S « 

co o 

g » 03 

°- 9-"b. 
O E .S ^ 



•s 

1 

I 





03 


X) 


i- 


? 


O 


X 

CO 


3 
O 


~a 


<U 


B 


o 


CO 


a 


♦-* 


X 



<u 



CD 

X 



(U 

■c 

J 

CU 
I 

x> 
H 

o 



I 

o 



I 

I 

I 



S 



B 

a. 



CO 

(U 

V- 

O 

U- 

■o 

B 

"^ 



■s -S 

2 •- 

x 2 

00 cu 



404 



i- 






o x 






is «m 


B 




8P« Ba 


3 




gg eT 


JD 






fe- 

3 




3 -O 
O V 


c 


4) 


u e 


3 


s 


8 £ 


O 

c 
O 


cS 


C5 P 


o 

c 

03 


a 
G 

a 




CO 


o 



« 


« 


ts 


« 


S 


1 


i 


1 


o 


O 


o 


o 


Ph 


fe 


Uh 


1*4 


TJ 


•g 


•g 


TD 


§ 


1 


§ 


g 


j? 


* 


« 


| 


£ 


£ 


£ 


£ 



co 


co 


« 


<l> 


E 


i- 


O 


o 


Uh 


Uh 


T3 


•o 


C 


c 


C8 


ca 



o 



<u 



£ £ 



la 
O " 



BO 



a 
m 

a^ 

3 "O 

o o 
k. e 

OS 

* E 



c 
8 

| 

3 

I 

o 
s 

es 



i 

CO 

c 
4> 
a 

O 







GO 


GO 


4> 


a> 


s 


E 


o 


o 


Uh 


u< 


•a 


T5 


c 


C 


CS 


C3 



V 



<u 



£ £ 



u 

tv 
en 

Q 

go 

ao 
C8 



C 

ca 



o 

■g S 

o S 

c a 

M Cj 

4J <L> 

a > 

CQ T3 



I 

c 
a 
O 



&o £ 

Ml 

£ K 



a 



Jg ca 

ffl E 



II 

^ o 

5 -3 

Em GO 

■81 

E w 

§> a" 

js a 
CQ o 



ja .£7,2 

oo o y 

.2 5 •§ 



co 

CO 

e 

a 

O 

"O 

c 
co 

J 

u 

u 

-E 

u 

1> 
<u 
ao 

O 



e3 

CJ CO 



CO 



s »•§ 

f .8 -a 

u . r c 

X> fi 3 
6 6 j? 

x> s a 
3 II 

>-4 co X 



T3 

0) 

3 

a 

s 
o 
o 

I 

I 

CN 

i 

1 



a 

08 

s 

e 
es 



ao 


es 


08 


X 




k> 


U 


D 


= 


T> 


o 


C 


*rf 


3 


08 


O 


*J 


E 


0J 

en 


o 


Sv 


p 


> 


fe- 
es 




CQ 



I 

I 
I 



CO 

CQ 



C 

a 
O 



IP -M 



I 
I 



a 



E 

3 

o 



CZ3 
i 

E 

3 

o 

I 

co 

co 

£ 

a 
U 



c 



c 

M 



co 

CO 

a 
>> 

U 
■ 

<U 

JB 
o 

o 
o 

OX) 

O 



CO 

I 

CQ 



405 



T3 

u 

3 

C 

- F— 

c 
o 
o 

I 
I 



x 

S3 

H 



!■ 














O -° 














«S «s 














sp« fi. 














B R 














S T3 


** 


** 


*-• 








CO 


x/i 


C/3 










| 


i 






<D 

c 


O 9i 


o 


o 


o 






»- e 


tLH 


[In 


u- 






a z 


-a 


•o 


T3 






£ 


Class 
Com 


c 


c 


C 




i 

On 


"O 


JS 


i2 


_rt 


x 


c 






6 


5 

X 
CO 


3 

"a. 
D 


u 














O " 














«JS *-. 














ffi, — 














gB SB 




























3 "3 














c a> 


+■» 




■«-» 








s_ e 


M 




CO 








Class Gi 

Combi. 





X) 


o 

E 

O 






u 

c 


"O 


3 
C 


T3 






E 


c 


X 


C 




CJ 


T3 


eg 


CO 


.5 


X 


c 




I 


1 

CO 




CO 


'3 

ES 

0-, 


est 

"a, 
D 








u 


"O 


u 




e 


X 

5 

x 

CO 

T3 

C 


X) 

2 

x 


3 
■•-» 
X 

I 


c 

00 


;c 

E 

D, 

CO 

3 

o 


La 

> 

o 

c 


*E 


J3 


co 


o 


_c 


1) 


"S. o 


y 

B 

Q 

en 
M 

es 


S3 

i 


x 
'1 


u 
IS 


p 

> 

o 


o 

es 

| 
G 

X 
u 
O 


03 « 

S | 


43 

'5 b 




T3 

C 


6 
II 

II 


§ & 


u 


co O 

Ou-O 
U 3 


_>> o 

^r co 

Xi T3 
o 3 
J 3 


CO 

o 
p 

> 


o 

'X 

S3 

3 
cr 

< 


1*3 

CO 00 


Ou 














es 














S 














T3 














a 














es 














■ 














09 














es 




























u 


Xi 








a> 




I 
o 

5 
6 


3 

1 

CO 
i 

co 
co 


X 


o 

CJ 

s 


X 
CO 


E 
U 


11 

c 


ox 
6 

> 


i 


CO 
1 


J. 

est 


1 

3 


c 

03 


>i 


co 


S3 


•c 


O" 


"5. 




U 


0Q 


o 


PQ 


< 


D 



CO 

a 

S3 



C 

6 

I 

6 

x 

co 

y 

co Cu 

S3 C3 

O £ 

— f- 

-2 § 

e § 

co CD 

i_ 00 

o u 

co "* 

C <N 

O v\ 

• 2 s 



a 
E 


X 


o 


X 


CJ 


-*-» 


u. 

O 


5 


■4— 


■o 


in 


u 


O. 
C3 

E 


C3 

a. 

3 



X ° 

6 co 

D. CD 

3 co 



oo- 

5J 



vO 



5 "S 

> .-3 

CO ^ 

U 

CO CO 

co a. 

C3 c3 



406 
during March 1962. Interpretation of the vegetation community types and distributions 
in these photographs was accomplished in 2 steps. The region included in the USGS 
topographic quadrangles around the sill (Pocket, Billy's Island, Craven's Hammock, 
Spooner) was included in the first set. The photographs included in each quad area were 
mosaicked and temporarily fixed to a mounting board. A georeferenced, mylar template 
was created for each quadrangle. The template for each quadrangle included any 
streams, rivers, and ditches recorded from the edited USGS 1994, 1:100,000 hydrologic 
feature DLG's, tramline features from the composite tramline coverage, refuge property 
and wilderness area boundaries digitized from refuge notations on USGS 1:24,000 1994 
topographic maps, locations of benchmarks installed for the GPS topographic survey 
(Chapter 2), and reference tic marks for matching the mylars among quad areas. Mylars 
were placed over the mosaicked photographs, matching hydrologic features and tramline 
and refuge boundary evidence where detectable on the photographs. A minimum 
mapping unit (MMU) of 5.76 ha (1 cm = 240 m x 240 m) was used to delineate areas of 
vegetation and land features into 6 categories: upland forest, wetland forest, shrub, 
prairie, open water, bare ground-urban. Boundaries of the vegetation communities were 
traced from the photographs onto the mylar. Photographs were repositioned as necessary 
to adjust for edge distortion. Areas smaller than the MMU were not delineated; the 
predominate vegetation type in the MMU was chosen to represent the location's 
vegetation type. Mylars were edge-matched and vegetation community boundaries 
transferred to the adjacent mylar where they were continuous between quad areas. After 
vegetation boundaries for the quad were traced, each polygon was given a polygon 



407 
number and vegetation label identified from the photo stereo pairs. Each mylar map of 
vegetation polygons was digitized into ARCINFO coverages using the tic marks for 
geographic reference. Digitized polygons were proofed for discontinuous arcs and 
missing or multiple labels. 

Comparison of the digitized mylars and the 1990 satellite image classification 
(see below) indicated that some distortion was present on several of the mylar polygon 
maps, most likely originating on the aerial photographs. This distortion needed 
correction so that comparisons made among maps would more likely indicate true 
vegetation changes, rather than changes due to these distortions. Locations were selected 
from the 1952 and 1990 coverages where similar features were discernable but location 
differed, indicating that change had not occurred along the vegetation polygon edge but 
the edge location was distorted, as well as where locations were not distorted. Select, 
undistorted points assured that fit remained where it already occurred. Multiple points 
were selected until the calculated transformation order indicated the a root mean square 
error term of < 100 m. The transformation was then applied to the 1952 coverage and 
the resultant transformed image was visually compared to the 1990 map to determine if 
the transformed image was a suitable match, or if additional points were necessary for 
calculating another transformation to achieve a better match. 

The second set of photo interpretation areas was randomly selected to represent 
regions of the swamp more distant from the sill. Poor photo quality prevented photo 
interpretation of the entire remainder of the swamp area; therefore, a subset area was 
randomly selected to represent pre-sill vegetation community distributions in regions 



408 
more distant from the sill. The entire swamp coverage was gridded into 5.76 ha cells and 
numbered with a unique X and Y combination. A random numbers table was used to 
select 2-digit numbers representing the X and Y values of the cells; the quad map within 
which the cell occurred became an area selected for photo interpretation. Cells were 
reselected until 4 separate quad areas were chosen (Double Lakes, Chesser Prairie, 
Strange Island, Waycross SE). Blackjack Island quadrangle area was originally selected 
and interpreted, but was later discarded due to extreme photo distortion, as was the lower 
1/8 of Chesser Prairie quadrangle. Photo interpretation, edge matching, and digitizing 
procedures followed those previously discussed. Transformations necessary to correct 
photo distortions in conversion to the arc coverages were calculated as indicated above. 

The final quad areas were joined into one coverage using the transformation 
matrix developed during edge matching, and gridded tolOmxlOm cells in ARCGRID. 
The "focal majority function" was used in ARCGRID to re-sample the grid to 240 m and 
320 m cells, to produce maps of the original interpretation resolution (240 m) and for 
comparison with the 1977 vegetation map (320 m) (see below). These coverages were 
used in change assessments discussed below. All comparisons were made with maps of 
each resolution and class combination to detect artifacts of scaling and vegetation species 
groupings into classes. The area interpreted from the 1952 photos covered 58% of the 
total refuge area (Figure 4-5). 



409 




a 

'C 

3 

C 

o 

§ 

> 

ai 



c 
_o 

ca 
2 

u 

o 

c 

ID 

O 

'-4- 

o 

*-^ 
3 

<u 

c 

■3 

I 



o 

43 
■ 






410 
17 Years With-Sill and 22 Years Post-Fire Vegetation (1977) 

A map created by McCaffrey and Hamilton (1982) represents Okefenokee 
Swamp vegetation 17 years after the Suwannee River Sill was constructed and 22 years 
after the 1954-1955 wildfires. This map was created by mosaicking 1:30,000 color- 
infrared aerial photographs recorded during November 1977, and interpreting the 
vegetation communities with 10 ha MMUs (Hamilton 1982). Vegetation classes used in 
this interpretation and the re-groupings used for comparisons with the pre-logging, 1952, 
and 1990 vegetation maps in the current study are listed in Table 4-2. This map included 
all of the area within the refuge boundary. 

The paper map provided by McCaffrey and Hamilton (1982) was converted to 
digital form (1 pixel=7. 15 m) on a flatbed scanner. Extraneous detail was removed and 
polygon labels representing vegetation types added using ARCEDIT. Transformation of 
the scanned map was necessary to correct distortion probably originating in the 
unrectified aerial photos used to make the original map, and to reference it to a 
coordinate system (NAD27 UTM zone 17) common with the other vegetation maps used 
in this study. Points along polygon edges were matched to common features discernable 
in the 10 m resolution merged panchromatic and multispectral 1990 SPOT satellite 
image (see below); the common feature edges were interpreted to be unchanged during 
the interval. Adjustments were made throughout the scanned 1977 coverage to match the 
map polygons to correct locations on the registered image, and areas already in 



411 
agreement were not modified. The final transformed map was gridded to 320 m cells in 
ARCGRID for comparison with the other vegetation maps (Figure 4-6). 
30 Years With-Sill and 35 Years Post-Fire (1990) 

Swamp vegetation community types and distributions 30 years after sill 
construction and 35 years after the 1954-1955 wildfires were represented by the 
vegetation map created from the merged panchromatic and multispectral 1990 SPOT 
satellite imagery discussed in chapter 2 (Figure 4-7). The 10 m resolution vegetation 
map was re-sampled using focal majority to 240 m and 320 m grid cells in ARCGRID for 
comparison with the pre-logging, 1952, and 1977 vegetation maps, and vegetation classes 
were re-grouped as indicated in Table 4-2. 
Wildfire Burn Area Maps 

Areas of the swamp burned by wildfires during 1855-1993 were digitized to 
provide fire polygons to compare with vegetation, logging, and hydrologic feature maps. 
Estimates of areas burned by wildfires during 1855-1937 were summarized from Trowell 
(1987); area burned during 1938-1993 was summarized from refuge records. Procedures 
and data used to develop these maps are discussed in Chapter 5. These fire polygon 
maps were combined into fire sets (Table 4-3) for comparison with vegetation changes 
occurring during various intervals, by intersecting fire polygons and dissolving the 
common borders in ARCEDIT. Overlapping polygons and common borders were 
dissolved to create maps representing total burn coverage during each interval. These 
maps were used to determine vegetation occurring prior to fires that subsequently 



412 



i 

s h 



w 



+ 



00 




a. 
E 

a 

u 
1) 

o 

c 

O 
o 

o 

oo 

o> 



a 

o 



E 
a 

x 

-a 

a 
a 

>> 

<u 

IE 

IB 

Q 



JQ 

T3 
fi 

a 

u 

O 

a, 

03 

£ 
o 

c 
o 



> 

-o 



t/5 

c 

T3 ON 

C 

H 

Q 

00 



60 

c 



NO 
I 

i 



c 
o 

1 

> 



413 





«&» 







2 

R 

% 



■IP ° 

& 1 






00 

I 

u 



H 

O 

Oh 

o 
o 

On 



C 

o 

ca 
u 
(C 

'(7! 
o> 
js 

"o 

CO 

c 
-o 

U 

«J 
9 

"u 

T3 

en 

C 

o 

i 
.£5 



C 

o 

1 

I 

■ 

| 

- 



414 



Table 4-3. Date groupings for wildfire map sets and for vegetation distribution 
comparisons. 







Data Sources for 


Purpose of Selected 


Fire Set 


Interval of Years 


Wildfire Coverage 


Time Interval 


A 


1855-February 


digitized polygons 


includes wildfires occurring from 




1952 


(1855-1951), point 


pre-logging to pre- 1952 aerial 






locations with radius 


photography 






buffers* (1939-1951) 




B 


March 1952- 


digitized polygons 


includes wildfires occurring after 




October 1977 


(1952-1968), point 


the 1952 and before the 1977 






locations with radius 


photography 






buffers (1952-1977) 




C 


November 1988- 


digitized polygons 


includes wildfires occurring after 




11 May 1990 


(1978-1989), point 


the 1977 photography and before 






locations with radius 


the 1990 imagery 






buffers (1978-1989) 




D 


March 1952- 


digitized polygons 


includes wildfires occurring after 




December 1955 


(1952-1955), point 


the 1952 photography until the end 






locations with radius 


of the extensive 1955 fires 






buffers (1952-1955) 




E 


November 1977- 


digitized polygons 


includes wildfires occurring soon 




December 1980 


(1978-1979), point 
locations with radius 
buffers (1978-1980) 


after the 1977 photography 


F 


12-May 1990- 


digitized polygons 


includes wildfires occurring soon 




1993 


(1990-1993), point 
locations with radius 
buffers (1990-1993) 


after the 1990 imagery 



a "Radius buffers" refers to an estimated burn area. Fires that had size estimates and 
location in the refuge records, but no location map were plotted as a point. A 
surrounding circle estimating the fire area was also plotted to roughly approximate the 
burned area. These areas are referred to as a "radius buffer" around the approximate 
location of the fire origin. 



415 
burned, vegetation regrowth following fires, areas and frequency of reburning, vegetation 
changes occurring with and without fire, and coincidence of logging and wildfires. 
Comparisons are detailed in Chapter 5. 
Map Comparisons 

All comparisons among maps were conducted in IMAGINE using the matrix 
procedure for creating new maps of changed, burned, or logged and burned areas, or the 
report summary procedure for generating reports for each comparison. A common 
vegetation classification for all maps was used for these comparisons (Table 4-2). 

Results 

Overall Changes in Vegetation Distributions and Composition 

Throughout the nearly 150 years examined in this study there were natural and 
man-made, direct and indirect, and short-term and continuous disturbances in the swamp 
environment, processes, and vegetation distributions. Although only a short period of 
time (1850-1993) was chosen for observation, evidence of many disturbances is present, 
suggesting the severity of the disturbances and the variance of the system's resilience to 
different disturbance types. 

During the pre-logging surveys of 1850-1890, swamp explorers saw an ecosystem 
similar in many ways to that existing today. The predominant community types in the 
swamp during the 40 years prior to intensive logging (1850-1890) were cypress-gum- 



416 
shrub (34.8%), Smilax spp.-shrub (26.1%, including aquatic and herbaceous prairies), 
and gum-bay-cypress-shrub (15.3%) (Table 4-4). In 1990 these vegetation types also 
covered over half of the swamp (24.9% cypress-gum-shrub, 17.9% Smilax spp.-shrub 
including shrub and prairie types, and 13.0% gum-bay-cypress-shrub), with a greater 
amount of bay communities (12.1% loblolly bay, 18.8% bay-shrub) occurring in 1990 
than before logging (1.1%) (Table 4-5). Total forested area was nearly equal during the 
period of pre-logging surveys (1855-1890) and 1990 (67.4% and 62.8%, respectively), 
suggesting that by 1990 there may have been recovery of the pre-logging total landscape 
structure, although not necessarily by the same species or in the same locations (see 
below). During the intervening period the swamp landscape was more heavily covered 
by shrubs. In the 58% of the refuge area examined on 1952 aerial photographs, shrub 
(39.7%) and wet forest (39.5%) areas were nearly equal in total coverage, with prairie 
(13.0%) and upland pine (7.4%) comprising the remainder (Table 4-6). By 1977, 48.7% 
of the area in wet forest in 1952 had changed to shrub, probably a result of the 1954-1955 
fires (see below), with total wet forest coverage of 28.2% and shrub coverage of 54.5% 
(Table 4-7). Shrub and wet forest types were varied in composition in 1977, with no 
forest type > 1 1.0%, and shrub types <15.0% overall cover. By 1990, wet forest (57.4%) 
and shrub (28.9%) coverage was nearly equal to pre-logging proportions (63.0% wet 
forest, 32.6% shrub). Wet forest composition was predominantly cypress-gum-shrub 
(24.9%), gum-bay-cypress-shrub (13.0%), and loblolly bay (12.1%) in 1990; shrub areas 
were dominated by bay-shrub (18.8%) and a shrub mixture (7.0%). Upland pine 



417 



Table 4-4. Okefenokee Swamp vegetation composition estimated from pre-logging 
surveys conducted during 1850-1890. 



Vegetation* 

Class on 

Original 

Map 


Area (ha) 


% of Total 
Area 


Vegetation 

Class on 

Grouped 

Map 


Area 
(ha) 


%of 
Total 
Area 


Gum-Maple- 
Bays 


826 


0.5 


Wetland 
Forest 


99628 


63.0 


Gum-Bay- 

Cypress- 

Shrubs 


24200 


15.3 


Upland Pine 


6971 


4.4 


Cypress- 
Gum-Shrubs 


55032 


34.8 


Shrubs 


51515 


32.6 


Wetland Pine 


15601 


9.9 


Prairie 


141 


0.1 


Oak-Hickory 


2021 


1.3 


Bare Ground- 
Urban 


n/a b 


n/a 


Ogeechee- 
Cypress 


139 


0.1 


Open Water 


n/a 


n/a 


Bays 


1766 


1.1 








Cypress- 
Shrubs 














Bay-Shrubs 


43 


0.03 








Pine- 
Palmetto 


6971 


4.4 








Smilax- 
Shrubs 


51515 


26.0 








Carex- 
Nymphaea 


141 


0.1 








Cypress 
Classes 


55171 


34.9 








Gum-Bay 
Classes 


26835 


17.0 









Table 4-4--continued 



418 



Vegetation* 

Class on 

Original 

Map 


Area (ha) 


% of Total 
Area 


Vegetation 

Class on 

Grouped 

Map 


Area 
(ha) 


%of 
Total 
Area 


Bay with 
Shrubs 

Shrub-Prairie 


1809 
51656 


1.1 
32.6 









"Vegetation classes are listed individually and as grouped for the 6-class map. 
b Bare Ground-Urban and Open Water classes were not represented in the pre-logging 
survey notes and if they existed at the time, are assumed to be included in the remaining 
4 classes. 



419 

Table 4-5. Okefenokee Swamp vegetation composition estimated from an 1 1 May 1990 
SPOT satellite image. 



Vegetation Map and Classes 


Area (ha) 


% of Total Area 


21-Class Vegetation Map 






Mixed Wetland Pine 


4422 


2.8 


Loblolly Bay 


19357 


12.1 


Ogeechee-Cypress 


66 


0.04 


Gum-Maple-Bays 


4254 


2.6 


Pine-Cypress-Hardwoods 


3167 


2.0 


Gum-Bay-Cypress-Shrubs 


20949 


13.0 


Cypress-Gum-Shrubs 


40023 


24.9 


Upland Pine 


172 


0.1 


Clearcut-Sparse Pine 


18 


0.01 


Dense Pine 


5207 


3.2 


Sparse Pine 


3380 


2.1 


Shrubs 


11295 


7.0 


Briar-Shrubs 


4500 


2.8 


Mixed Upland- Wetland 
Shrubs 


536 


0.3 


Bay-Shrubs 


30250 


18.8 


Sedges-Ferns- Water Lilies 


10962 


6.8 


Aquatic Grasses 


105 


0.1 


Water Lilies 


1500 


0.9 


Bare Ground-Urban 


307 


0.2 


Agriculture-Lawn 


7 


0.004 


Open Water 


78 


0.1 



Table 4-5-continued. 



420 



Vegetation Map and Classes 


Area (ha) 


% of Total Area 


Cypress Classes 


40089 


25.0 


Gum-Bay Classes 


74810 


46.6 


Bay with Shrubs 


49607 


30.9 


Shrub-Prairie 


28362 


17.7 


9-Class Vegetation Map 






Wetland Forest 


30308 


19.5 


Gum-Bay-Cypress-Shrubs 


20872 


13.0 


Cypress-Gum-Shrubs 


39978 


24.9 


Upland Pine 


8670 


5.4 


Shrubs 


16056 


10.1 


Bay-Shrubs 


30184 


18.8 


Prairie 


977 


7.8 


Bare Ground-Urban 


321 


0.2 


Open Water 


80 


0.1 


6-Class Vegetation Map 






Wetland Forest 


92159 


57.4 


Upland Pine 


8670 


5.4 


Shrubs 


46400 


28.9 


Prairie 


12523 


7.8 


Bare Ground-Urban 


321 


0.2 


Open Water 


80 


0.1 



421 



Table 4-6. Okefenokee Swamp vegetation composition estimated from 1952 black and 
white aerial photography. 







% of Total Area 


% of Total Area 


Vegetation Class 


Area (ha)" 


(240 m MMU) b 


(320m MMU) 


Wetland Forest 


36941 


39.5 


40.9 


Upland Pine 


6889 


7.4 


4.8 


Shrubs 


37124 


39.7 


41.2 


Prairie 


12138 


13.0 


12.9 


Bare Ground-Urban 


311 


0.3 


0.2 


Open Water 


117 


0.001 


0.06 



a Approximately 58% of the refuge is included in the interpreted area. 
b Photographs were interpreted with a minimum mapping unit (MMU) of 240 m. The 
interpreted map was re-sampled to 320 m cells to compare with the 1977 vegetation map. 
The proportions resulting from this re-sampling are listed in the last column. 



422 



Table 4-7. Estimated Okefenokee Swamp vegetation composition compiled from 1977 
color-infrared photography interpreted by McCaffery and Hamilton (1982) with a 
minimum mapping unit of 320 m. 



Vegetation Map and Classes 



Area (ha) 



% of Total Area 



Needle-leaved Evergreen 
(Wetland Pine) 


446 


Mixed Wedand Pine 


2304 


Scrub Pine 


1829 


Bay 


2027 


Bay-Cypress 


1525 


Mixed-Cypress 


5567 


Scrub 


16700 


Cypress 


3196 


Blackgum 


7821 


Scrub-Prairie 


2790 


Upland Pine 


9741 


Shrub 


21124 


Scrub/Shrub 


14988 


Shrub-Pine 


2041 


Shrub-Prairie 


20737 


Shrub-Bay 


11952 


Shrub-Cypress 


6469 


Cypress-Shrub-Prairie 


8489 


Herbaceous Prairie 


4171 


Aquatic Prairie 


13540 


Bare Ground-Urban 


n/a a 


Open Water 


4 


Cypress Classes 


24167 



0.3 

1.5 

1.2 

1.3 

1.0 

3.5 

10.6 

2.0 

5.0 

1.8 

6.2 

13.4 

9.5 

1.3 

13.2 

7.6 

4.1 

5.4 

2.6 

8.6 

n/a 

0.003 

15.3 



Table 4-7-continued 



423 



Vegetation Map and Classes 


Area (ha) 


% of Total Area 


Gum-Bay Classes 


24521 


15.6 


Bay with Shrubs 


15504 


9.8 


Shrub-Prairie 


61613 


39.1 


10-Class Vegetation Map 






Wetland Forest 


41570 


26.4 


Upland Pine 


9763 


6.2 


Shrubs 


58890 


37.4 


Shrub-Bay 


11967 


7.6 


Shrub-Cypress 


645 


4.1 


Cypress-Shrub-Prairie 


8503 


5.4 


Scrub-Prairie 


2834 


1.8 


Prairie 


17636 


11.2 


Bare Ground-Urban 


n/a 


n/a 


Open Water 


4 


0.003 


6-Class Vegetation Map 






Wetland Forest 


44404 


28.2 


Upland Pine 


9763 


6.2 


Shrubs 


85816 


54.5 


Prairie 


17636 


11.2 


Bare Ground-Urban 


n/a 


n/a 


Open Water 


4 


0.003 



" Bare Ground-Urban class was not represented in the 1977 vegetation map; this class is 
assumed to be included in the remaining classes. 



424 
coverage remained fairly constant since the initial surveys (1850-1890: 4.4%; 1952: 4.8- 
7.4%; 1977: 6.2%; 1990: 5.4%), while prairie coverage gradually declined (1952: 13.0%; 
1977: 1 1.2%; 1990: 7.8%), and shifted from aquatic to herbaceous prairie type during 
1977 to 1990. 

Vegetation composition has not changed uniformly across the landscape over 
time. Some areas and vegetation types have been more constant in composition than 
others. Distribution of upland pine communities has remained fairly constant since the 
pre-logging period, despite the effects of fire and logging (see below). Areas in 
persistent upland pine (i.e., were occupied by upland pine at the start and end of the 
interval) during 1890-1952 (63.8%), 1952-1977 (71.6%), and 1977-1990 (62.2%) have 
been intermittently replaced by wet forest communities (1850-1952: 18.5%; 1952-1977: 
21.6%; 1977-1990: 26.7%) (Tables 4-8, 4-9, 4-10). Prairie, shrub, and wet forest 
community distributions have shown less constancy. Areas remaining in prairie 
vegetation have declined from 78.1% during 1850-1952, to 57.9% during 1952-1977, and 
to 28.7% during 1977-1990. Replacement has primarily been with shrubs (1952-1977: 
35.8%; 1977-1990: 40.1%), although some change to wet forest has also occurred (1850- 
1952:16.9%; 1952-1977: 5.7%; 1977-1990: 30.9%). Area of persistent wet forest 
distribution has increased since the pre-logging period (1850-1952: 43.5%; 1952-1977: 
45.4%; 1977-1990: 80.0%), while persistent shrub coverage has fluctuated (1850-1952: 
38.7%; 1952-1977: 64.5%; 1977-1990: 34.3%). The increase in wet forest coverage 
during 1977-1990 was due to areas in shrubs during 1952-1977 changing to wet forest 
during 1977-1990. Most of this change from shrub coverage was to cypress-gum-shrub 



425 



B 


(N 


3 


m 


bfi 


On 


c 


"~" 


O. 


C 


a. 
E 


i/3 


E 

3 

E 


u 

•a 



1 1 

• 4) 

On c 



© S 



. M 

00 fc 

3 
C U 

•c o 

J ° 

g a 

« .E 
^ a 

GO 52 

(U ^ 

OJ u 

g.i 

O ff 

.s I 

tjn -C 

■ 5 <*- 
t o 

3 no 

O i- 

<A « 

oo i> 

C 3 

J-g 

- T3 

•2 c 

i a 

> 

> o 
.22 * 

u s 

co O 

•^ CO 

S ' c 

_) a 

. E 
oo o 

<* I 

e2«S 






4> 

a 
O 



■a itj 

C On 

S p* 

£.s 

O m 

<U OB 
b. X! 

OB k. 

00 3 



4) «S 

H — 



s 
- 
— 



u 

a 
£ «s 

"O On 

a «-< 

es 



I 

E 

— m 

B •" 

ee = 

~ "™ 
6 



S 9 
.2 c B 

■**•—■ 00 

£^2 



CM 

© 



CM 

o 



no 
© 



o 
o 



SO o 

o — ; 



On 



© oo 



00 

no 



o 

<N 



On 



NO 

ON 



00 



On 

O 
On 



ON 

cm 



CM 



© 



o 


— 


o 


o 
rn 


</"> 


CM 



oo 
m 



ON 



CM 






NO 
CM 



rn 



NO 



00 

rn 

NO 



CM 



CM 
© 



ON 



SO 

V") 



ON 



-h ON 

CnJ NO 



■* 


in 


r* 


NO 


00 


CM 



«0 rj- 

— i Cn) 



Cn| 



CO 



| 



& 



^^ 


o 


5 t» 


(/> 


CO w 


<Z1 


i <u 


a> 


g | 




3 >. 


>^ 


a u 


U 





;> 




c 




a, 


a 


"O 


>' ■«: 


c 


If 




££ 



o 

a 

o 

I 

c 



o 
J* 

o 



CO 

O 



i 

JS a 

o 8 

00 >, 

o u 



oo 





| 






5 






i 

CO 

(A 


X) 




U 


oo 


m 






>^ 


O. 


>^ 


C3 


>-. 


f3 


pa 


U 


CO 

















426 


<u 
















£ 
















<-» 
















k. 
















a 






























+-» 


4 J 

*» s ■ r- 














1 

D. . 
a-f- 


Bai 

Grou 

Urbai 

197 


© 


o 


o 


o 


O 


o 


re r» 
















S £ 






























£ a 


L. t» 














3 — 

C CO 

C co 


ere — 


o 


o 


o 


fN 

o 


o 


© 


'£ J 


0£ B 








o 




St3 
















CD 
















H S 


0> I-- 














1952-197 
the spec 


Prairi 
in 197 




Os 
O 


SO 


Os 


o 


48.6 
















during 
Hiring in 


Scrub- 
Prairie 
in 1977 


o 


CN 

© 


ro 


© 


Os 

cn 


o 


wamp 
2 occi 
















K ■£ T h- 














okee S 
in 195 


Cyprc 
Shru 
Praii 
in 19 


SO 
OS 


o 




SO 


Os 

so 


o 


ing in Okefen 
getation class 
















A i£ 














Shru 
Cypr 
in 19 


SO 


©' 


o 


00 


o 


o 
















occurr 
f the ve 


^.5^ 














Shru 
Bay 
197 


Os 

SO 




so 


1*^ 


p 


o 


8 o 

DOxP 
C o^ 
















x> r- 
















Shru 
in 19*3 


SO 


irj 


— 


fN 


</-> 


rn 


ioncl 
uesa 




r*i 


Os 

m 


sd 


f»' 


«-> 


re re 
















#z > 


•a s 














o> 


lip: 














ox) -a 


** 


* 


r*- 


© 


<r> 




vel v 
eport 


Weti 

Fore 

19 


CO 


rs) 




iri 


</-> 


o 


ndscape-le 
320 m. R 
















Upland 
Forest 
in 1977 


SO 


so 

r*- 




so 
© 


en 


o 


a-j 






























o> £ 


e 


TD 








-!. 




Table 4- 
compari 


i« 


§ « 


c 


x> 


_<U 


"O 

c c 


r- Ir: 


w 2 
G 


Wetl 
Fore; 


Upla 
Pine 


2 

00 


'C 
're 

B 


Bare 
Grou 
Urba 


II 

















£ 



All 



oo 

c 

1 

o. 

CO o 

[I 

3 C 

£ 

3 vi 

.£ « 



ON S 

On o 

i» 

oo c 
c •- 
C oo 

3 C 

•6 ■£ 

& p 

£ t-> 

eo o 

\S 

on £■ 

<L> OS 

? 8 

61 

c o 



t5> 



(so ra 

.£ "S 

§ g 

o <u 

O X! 

oo ° 

C nO 

eO s 

.2 8 
§ J 

oo ^ 

U T3 
> (D 

V, € 

> a 

g if 

I " 

CO e 

o H 

9 r^ 

T3 ° 

c £J 

a <^> 
2 § 

;1 



X> 
CO 
H 



CO 

a 

E 

o 



i B 




















Bare 

round 
rban i 
1990* 


00 

© 


d 


CM 

o 
o 
o 


CO 

o 
o 


o 


o 


o 


© 
o 


o 






d 


d 








d 




o s 


















B 








































Open 

ater 

1990 


o 


o 


d 


o 


o 


o 


o 


CO 

d 


o 


£ 




















01 o 






















■*r 


ON 


r- 


<»o 


Tf 


_ 


© 


r~ 


•<3- 


M — 

U _ 


©' 


CM 


00° 


d 


CM 


ON 


wo 


00 
CM 


cm' 

CM 


O. = 




















G» © 




















■ X! OS 

£> 3 On 


ON 


-; 


o 


</o 


■** 


Tt 


ON 


r~; 




« i- — 


Tf 


co 


00 


r--' 


On 


CO 


CM 


co° 


o 


K — - 


— 


cm 


CM 


rf 


CM 


CM 




00 .S 




















i B 




















W) , — 




















8 s 2 9 


_ 


tn 


NO 


ON 


CM 


00 


00 


00 


NO 


O. 3 = <* 


r~' 


CO 


«o' 


d 


<ri 


d 


d 


CM 


r-' 


CM 


CO 


CM 


CO 


CO 


•* 


CM 


r* 


Gum-Bay- 

Cypress- 

Shrubs in 

1990 




















r- 


ON 


00 


CO 


00 


00 


r- 


r^ 




co 


r-' 


On 


r-' 

CM 


On 


V£> 


zz 


Tf' 


o 




















B 








































■s® 


_ 


no 


VO 


NO 


<T) 


r^ 


r~- 


■t 




2 g 


«/"> 


— i 


On" 


d 


CO 


NO 


to 


NO 


o 


00 




















"O E 




















etlan 
rest i 
1990 


ON 


00 


CO 


CO 


<n 


— t 


NO 


"3; 




vi 


00 
CO 


r^ 


CM - 


ON 


CO 


oo 


CO 


o 


*fi 




















■O B 




















plan 
ine i 
1990 


CM 
CM 

NO 


i/o 

CM 


© 


00 

d 


CO 

d 


CM 

d 


Tt' 


s 

d 


o 


- - 




















Class in 

1977 


Upland 
Pine 


Wetland 
Forest 


in 

X 

5 

00 


■ 

X 

2 

X 

&o 


Shrub- 
Cypress 


Cypress- 

Shrub- 

Prairie 


Scrub- 
Prairie 


u 

'3 

E 

0. 


Open 
Water 



00 

c 
'E 

'CO 

E 
g 

u 



T3 
<U 
T3 

_3 

c 
x> 



T3 



3 
<A 

VI 

CO 



J 

V> 

'M 

c£ 

CO 

E 

c 
_o 

-♦-» 
CO 

00 

> 

ON 



1) 

— 



T3 
<U 

*-• 
C 

u 

CA 

u 
O. 



o 

c 

Vi 

J 

o 

c 

CO 

X) 



-o 

c 

3 

I . 

CJ CJ 

CO (O 
CQ 3 



428 
(23.3%) and gum-bay-cypress-shrub (17.9%) wet forest types in areas of shrub, shrub- 
bay, shrub-cypress, and cypress-shrub-prairie vegetation types in 1977 (Table 4-9). 
Logging Impacts 

Areas logged during 1890-1942 contained wet forest (88.2%), shrubland (4.3%), 

upland pine (7.5%), and prairie (0. 1%) before logging occurred (Table 4-11). By 1952 

the proportions of wet forest (41.6%), shrub (45.7%), and prairie (4.8%) communities 

had changed in the logged areas from pre-logging amounts, while upland pine remained 

fairly constant (7.3%) (Table 4-12). These coverages remained almost unchanged in 

1977 (Table 4-13). By 1990 wet forest coverage in logging tramlines had increased to 

69.5%, and shrub coverage had decreased to 18.6%; prairie (3.8%) and upland pine 

(7.9%) remained nearly constant (Table 4-14). Although the total coverage of wet forest 

in the logged areas had increased by 1990 to levels similar to pre-logging (1850-1890), 

the proportions of forest types differed. A mixture of cypress-gum-shrub (2 1 .4%), 

loblolly bay (20.6%), and gum-bay-cypress-shrub (15.6%) replaced the areas dominated 

before logging by cypress-gum-shrub (56.1%), gum-bay-cypress-shrub (18.4%), and wet 

V 
pine (7.9%). 

Vegetation changes occurring during 1952-1990 in the logged areas were similar 

to those occurring in the swamp overall (Table 4-15). During 1952-1977 in previously 

logged areas, persistent upland pine (85.7%) and persistent wet forest (58.0%) were in 

slightly higher proportions than in the swamp as a whole, whereas persistent shrubland 

(48.7%) and persistent prairie (41.8%) were lower (Table 4-16). This trend continued 



429 



Table 4-11. Estimated composition of logging tramline areas before logging occurred, 
recorded in surveys conducted during 1850-1890. 



Vegetation Class 


Area (ha) 


% of Total Area 


Gum-Maple-Bays 


150 


0.5 


Gum-Bay-Cypress-Shrubs 


6013 


18.4 


Cypress-Gum-Shrubs 


18335 


56.1 


Carex-Nymphaea 


43 


0.1 


Wetland Pine 


2582 


7.9 


Pine-Palmetto 


2451 


7.5 


Oak-Hickory-MagnoZ/a 


654 


2.0 


Ogeechee-Cypress 


72 


0.2 


S/w/ax-Shrubs 


1405 


4.3 


Bays 


1013 


3.1 


Cypress-Shrub 








Bay-Shrub 








6-Class Map* 






Wetland Forest 


28826 


88.2 


Shrubs 


1405 


4.3 


Upland Pine 


2451 


7.5 


Prairie 


43 


0.1 



a Bare Ground-Urban and Open Water were not distinguished from the other class types 
in this map. 






i 



430 



Table 4-12. Estimated composition during 1952 of areas previously logged. 



Vegetation Class 


Area (ha) 


% of Total Area 


Wetland Forest 


9357 


41.6 


Shrubs 


10268 


45.7 


Prairie 


1072 


4.8 


Upland Pine 


1629.6 


7.3 


Bare Ground-Urban 


127.9 


0.6 


Open Water 


14.1 


0.1 





■ 

431 




Table 4-13. Estimated composition during 1977 of areas previously logged. 


Vegetation Class 


Area (ha) 


% of Total Area 


Upland Pine 


3096 


9.7 




Needle-Leaved Evergreen (Wetland 
Pine) 


27 


0.1 




Bay 


856 


2.7 




Cypress 


696 


2.2 




Blackgum 


4038 


12.6 




Bay-Cypress 


99 


0.3 




Mixed Cypress 


1440 


4.5 




Cypress-Shrub-Prairie 


894 


2.8 




Mixed Pine 


193 


0.6 




Herbaceous Prairie 


271 


0.9 




Aquatic Prairie 


895 


2.8 




Shrubs 


3012 


9.4 




Scrub 


4080 


12.7 




Scrub/Shrub 


3570 


11.1 




Shrub-Pine 


317 


1.0 




Shrub-Cypress 


1082 


3.4 




Shrub-Bay 


4695 


14.7 




Shrub-Prairie 


2011 


6.3 




Scrub-Pine 


148 


0.5 




Scrub-Prairie 


617.3 


1.9 




6-Class Map* 








Wetland Forest 


12194 


38.1 




Shrubs 


15581 


48.6 




Upland Pine 


3096 


9.7 




Prairie 


1166 


3.6 


s Bare Ground-Urban and Open Water classes were not included in the original map. 


n 





432 



Table 4-14. Estimated composition during 1990 of areas previously logged. 



Vegetation Class 


Area (ha) 


% of Total Area 


Upland Pine 


40.7 


0.1 


Dense Pine 


1356 


4.1 


Sparse Pine 


1205 


3.7 


Clearcut-Sparse Pine 


5 


0.02 


Ogeechee-Cypress 


7 


0.02 


Gum-Maple-Bays 


1833 


5.6 


Gum-Bay-Cypress-Shrubs 


5095 


15.6 


Mixed Wet Pine 


1217 


3.7 


Loblolly Bay 


6759 


20.6 


Pine-Cypress-Hardwoods 


851 


2.6 


Cypress-Gum-Shrubs 


7003 


21.4 


Bay-Shrubs 


4423 


13.5 


Briar-Shrubs 


331 


1.0 


Shrubs 


1111 


3.4 


Mixed Upland/Wetland 
Shrubs 


214 


0.7 


Water Lily 


90 


0.3 


Sedges-Ferns- Water Lilies 


1100 


3.4 


Aquatic Grasses 


15 


0.1 


Open Water 


5.7 


0.02 


Bare Ground-Urban 


105 


0.32 


Agriculture-Lawn 


2 


0.005 


6 Class Map 






Wetland Forest 


22765 


69.5 


Shrubs 


6079 


18.6 



Table 4-14-continued. 



433 



Vegetation Class 


Area (ha) 


% of Total Area 


Upland Pine 


2606 


8.0 


Prairie 


1205 


3.7 


Bare Ground-Urban 


106 


0.3 


Open Water 


6 


0.01 



434 

Table 4-15. Proportions of the entire swamp and logged areas that remained in persistent 
vegetation types between intervals, and the predominant type of replacement where 
changes occurred during 1952-1977 and 1977-1990. 



Vegetation Type 


% of Swamp 
in Type, 
1952-1977 


% of Logged 

Area in 

Type, 1952- 

1977 


% of Swamp 

in Type, 

1977-1990 


% of Logged 

Area in Type, 

1977-1990 


Persistent 
Vegetation Type 










Upland Pine 


76.6 


85.7 


62.2 


66.6 


Wetland Forest 


45.4 


58.0 


80.0 


86.0 


Shrubs 


64.5 


48.7 


34.3 


21.0 


Prairie 


57.9 


41.8 


28.7 


29.1 


Predominant 
Change Type 










Upland Pine 


21.7 

(Wetland 

Forest) 


9.2 

(Wetland 

Forest) 


26.7 

(Wetland 

Forest) 


24.7 

(Wetland 

Forest) 


Wetland Forest 


48.7 
(Shrubs) 


37.4 
(Shrubs) 


15.6 
(Shrubs) 


10.5 
(Shrubs) 


Shrubs 


27.0 

(Wetland 

Forest) 


32.5 

(Wetland 

Forest) 


41.2 

(Wetland 

Forest) 


71.2 

(Wetland 

Forest) 


Prairie 

■ 


35.8 
(Shrubs) 


43.0 
(Shrubs) 


40.1 
(Shrubs) 

30.9 

(Wetland 

Forest) 


36.6 

(Wetland 

Forest) 

34.0 
(Shrubs) 



00 

c 
a 

Q. . 

3 C 
• 5 t« 

.£ 8 
2*3 

5 I 

S? en 
o> 111 

22.S 

c •- 

C 00 
3 C 

— CN 
(/3 <0 

cd os 

<U — 
E 

« C 

c '" 

S; o 

— c 

11 

DO"*; 

_c o 



s 



a 


g 


o 


co 


o 


</i 


en 


y 


!P 


3 


00 




c 


03 


03 


> 


J= 




U 




c 


B 


.2 


o 


'■5 


CN 




en 


u 


C/J 


00 




1 


c 
o 




VI 






V© 


« 


*«H 


s. 


1 


E 


u 


o 




u 


■ 


u 


H 


J= 

— - 























435 


c JB 






















V £ 


©\ 




















Q. at 




© 


o 




o 


O 


o 


o 






©£ 


a 

• MM 


















1 

c 


a> r- 




















T3 


Prairi 
in 197 




cn 
d 


— ' 




CN 
CN 


00 

5- 


o 


o 




to be include 


A v 




















| 


Scru 
Prair 




© 


CN 




CN 


o 


o 


o 






£ 


















to 

1 






















CO 






















• *■ 






















n 


ress- 

ub- 

irie 




















J 


©\ 




r-» 




"3" 


t^- 


■T 






(A 




B 


o 


ci 




CN 


00 


Tf 


o 






i> 1 




















•J3 


Shru 
Cypr 


■ 


1— 1 

© 


00 




CN 


>o 


o 


o 




1 


a = 

3 ^ 




Tf 


«/-> 




<r> 


>o 


TT 






o 


GO ■ 


0\ 




Os 




CN 


o 


CN 


o 




c 


OB t"- 




















1 


3 OS 




"•J; 


Tl" 




CN 


— 


CN 


o 




c 
B 






co 


e> 




CN 

cn 


CN 


00° 


2 






t/i .2 




















& 


rested 
etiand 




tN 


cn 




t— H 


00 


o 






1 


— 


ON 






© 

CI 


On' 


<n 


o 




% 


££ 


a 


















P* 


* — 


















52 






















| 
























■o e 




















o 


B — 
03 O 




r- 


>o 




>n 


W^ 


OS 






a 


"5..S 

- 6- 


Os 


oo 


cn 




CN 


CN 


OS 


o 




round-Urb 
ig classes. 


£ 




•a 




o 


w 


«U 


i 

*o « 






£ <N 




B 

es cu 


8 


9 


3 


'u 


e e 
i» 5 5 


B 




91 


02 






i- 
o 




c 

x: 


'5 

E 

a. 


S- O .£ 

es ■- C 

ee w - 


4* 




OS 


"Bare 
remai 



436 
during 1977-1990; by 1990 persistent prairie, wet forest, and upland pine occurred in 
proportions similar between logged areas and the swamp overall, while persistent shrubs 
were less abundant in logged areas (21.0%) (Table 4-17). The types of changes 
occurring in the logged areas were similar to those occurring throughout the swamp 
during these intervals. Prairie replacement during 1952-1977 and 1977-1990 was 
primarily by shrub (1952-1977: logged 43.0%, overall 35.8%; 1977-1990: logged 34.0%, 
overall 40.1%). Prairie was also replaced with wet forest, although less during 1952- 
1977 (overall 5.7%, logged 12.8%) than during 1977-1990 (overall 30.9%, logged 
36.6%). Upland pine was more frequently replaced by wet forest in the swamp overall 
(21.7%) than logged (9.2%) areas during 1952-1977, and replaced nearly equally by wet 
forest during 1977-1990 in the swamp overall (26.7%) and logged (24.7%) areas. Wet 
forest replacement by shrubs has decreased since 1952. During 1952-1977 wet forest 
was replaced by shrubs less frequently in logged (37.4%) than the swamp overall 
(48.7%); and, during 1977-1990 wet forest was replaced in lower proportions by shrubs 
in logged areas (10.5%) than by shrubs elsewhere (15.6%). 
Fire and Vegetation Change 

Effects of fire on Okefenokee Swamp vegetation distribution and composition in 
the landscape are detailed in Chapter 5. A summary of the swamp's response to 
wildfires is provided here. 

Prior to 1952 most wildfires in the swamp occurred in wet forest (61.0%; value 
represents the area of this vegetation type that burned during the specified interval), 



437 



I 



00 
C 



IS 

§ c 



£ 



o 5 

T 18 

|I 

••■ en 
oo (u 

'i 5 

3 C 

•o — 

-r- 00 

<u .5 
oo a 

.2 5 

£ 8 

g t- 

© .£ 
£ 9 

ON co 

•7 JS 
b c 

60 a 



a o 

T3 > 

c x; 



3 
U 

o 

o 

en 

u 
00 

c 

« 

o 

c 
o 



1) 
<S 
en 
(U 
_3 

"3 
> 



1 JS 

OJ o 
00 <N 

£> 1/5 

Dig 



JO 



B 

Cl 

E 
o 
o 



= - 

© = S 

9 — = 

M aw 

■ 
CO 



III 



k s 



5 1 & 

1*3 



00 -^ 



■ o 

E i 5 £ 



d 






o § o 
o ,-; 



tN 



O 

d 



VO 






d 



oo 



CO 



5 



■izi .5 



111 
Iff 



c 



02 



vO 

vd 



o 
vb 



o 






oo — 

— i d 



d 



•*J-° 









o" — 



r-° 



CM 



0> 






00° 



o 



o 

VO 



r-° 






r» <o o\ 

— < rn 00 
— < <N — 



pi vo c5 

-H Tj — 



00 

d 



o 
d 



0\ 



VO 
00° 









rs 



00 



vO 






00 

d o 



fN 



00 



o 



d 



vo 



"O 

c 
_■ A. 



II 



■ 
OS 

3 3 



« an 

IS SfcS 2 « "3 



£ u 



u w a. 






C 4> 

a. C ? 



438 
shrub (34.6%), and upland pine (4.4%) vegetation, and these fires were primarily in 
cypress-gum-shrub (38.9%), gum-bay-cypress-shrub (14.5%), and Smilax spp. -shrub 
(34.6%) (Table 4-18). By 1952 these burned areas had been revegetated with greater 
proportions of shrubs (41.3%) and prairie (14.0%), and less wet forest (39.7%) than 
before burning. Upland pine coverage remained persistent (4.5%). Prior to logging 
(1850-1890), 26% of the swamp surface fuel load (excluding peat) was in logging 
tramlines (Table 4-19); 95.0% of this logging tramline fuel was wet forest, and 89.0% of 
this was dominated by cypress (Table 4-20). During 1890-1942, 26.0% of the swamp 
was logged, and 23.0% of the area that burned during 1855-1952 was in logged areas 
(Table 4-21). Between 1890 and 1952, 64.2% of the logged area burned by wildfires. 
During 1952-1976 wildfires occurred in nearly all of the swamp, in vegetation 
types in proportion nearly equal to the overall swamp vegetation composition; vegetation 
that burned included wet forest (40.4%), shrub (39.1%), prairie (13.7%), and upland pine 
(6.3%) (Table 4-22). These vegetation types were replaced by shrubs (59.8%), wet forest 
(18.0%), prairie (16.3%), and upland pine (6.0%) by 1977. Most of the subsequent fires 
occurred in upland pine (56.5%), shrub (24. 1%), and wet forest (15.2%) communities 
(Table 4-23). By 1990 these burned areas had revegetated as upland pine (53.9%), wet 
forest (33.0%), shrub (1 1.9%), and prairie (1.1%). 
Vegetation Changes in the Areas Affected bv the Suwannee River Sill 

Two areas of the swamp have incurred hydrologic alterations since the sill was 
constructed. An area of 23,335 ha in the western and central swamp (Figure 3-9) is 



439 



Table 4-18. Vegetation types that burned after 1855 and before 1952, and the types of 
vegetation that occurred in the burned areas in 1952. 



Vegetation Type that 




Vegetation Type 




Burned After 1855 


Proportion 8 of 


Occurring in Burned 


Proportion of 


and Before 1952 


Sampled Area 


Areas by 1952 


Sampled Area 


Gum-Maple Bays 


0.6 


Bare Ground-Urban 


0.3 


Gum-Bay-Cypress- 


14.5 


Wet Forest 


39.7 


Shrub 








Cypress-Gum-Shrub 


38.9 


Open Water 


0.2 


Wetland Pine 


4.3 


Prairie 


14.0 


(Pond and Slash 








Pines) 








Pine-Palmetto 


4.4 


Shrub 


41.3 


Oak-Hickory- 


1.6 


Upland Pine 


4.5 


Magnolia 








Smilax-Shrub 


34.6 






Bays 


1.1 






Bay-Shrub 


0.03 






Ogeechee-Cypress 









Cypress-Shrub 










8 Proportions are of the area sampled, not necessarily for the entire swamp. Unburned 
area = 97287 ha, burned area = 87601 ha. 



440 



Table 4-19. Logging tramline fuel load estimates. 



Vegetation Type 


1855 (kg) 


1952 (kg)' 


1977 (kg) 


1990 (kg) 


Wet Forest 


1.8xl0 8 


7.9xl0 7 


4.6xl0 7 


1.0x10 s 


Shrub" 


3.8xl0 6 


2.8xl0 7 


5.4xl0 7 


1.7xl0 7 


Prairie 


9.5xl0 4 


2.4xl0 6 


2.6xl0 6 


2.7xl0 6 


Upland Pine 


5.8xl0 6 


3.8xl0 6 


7.3xl0 6 


6.1xl0 6 


Total Tramline 


1.9xl0 8 


l.lxlO 8 


l.lxlO 8 


1.3xl0 8 


Fuel 










Total Refuge Area 


7.5xl0 8 


4.6x1 8 


6.7xl0 8 


9.5xl0 8 


Fuel 











' Interpreted area includes 58% of the refuge; fuel volumes have been proportionally 

adjusted to compare with other sample periods. 

b Pre-logging shrub area includes some prairie; these types were not readily 

distinguishable in many of the survey descriptions. 

c Area logged in tramlines is 26% of total refuge area. 






441 

Table 4-20. Fuel load composition for fires occurring during 1855-1951, 1952-1976, and 
1977-1990. 





Fuel Load 








Vegetation 


Before Logging 


Fuel Load 


Fuel Load 


Fuel Load 


Type 


Began 


in 1952 


in 1977 


in 1990 




(kg/ha) 


(kg/ha) 


(kg/ha) 


(kg/ha) 


Prairie 


1.5xl0 6 


1.3xl0 8 


1.9xl0 8 


1.3xl0 8 


Shrub 


6.9xl0 8 


5.0xl0 8 


7.9xl0 8 


2.2xl0 8 


Wet Forest 


2.9X10 9 


1.5xl0 9 


l.lxlO 9 


2.7xl0 9 


(models 6 and 4) 










Upland Pine 


8.0xl0 7 


7.9xl0 7 


l.lxlO 8 


l.OxlO 8 


Cypress Only 


2.3xl0 9 


unknown" 


2.6xl0 8 


1.6xl0 9 


(model 4) 










% Wet Forest 


55.4 


unknown 


9.1 


32.7 


Area in Cypress 










% Wet Forest 


79.2 


unknown 


23.5 


59.8 


Fuel Load in 










Cypress 











a Cypress was not separated from other forested wetland species in the interpretation of 
the 1952 aerial photos. 



442 



tj — > 

0» S3 

m I ■= 

• Li — - 

O 2 08 

c b 



u — C 

< lZ 

«*■ 2 tj 



1 



OX) 

ox 



o — - = 

j5 e « J 



El « 

«- 1 9 P 



tj 

OX . « — 

OX ji £ 

— « u 

© 5 05 



_ TJ 
— I 

o 



3 <U 



§ 
s 

a 
s 

■ 

•s 






C * TJ 

is ° ■ 



oo 
as 



(N >r, m S 

^ oo oo £ 

p-i 55 «n S 

— ■ i oo >n 2 

O rn CO ^ 



oo 
oo 






SO 
00 



00 



CM 

00 



(N 



r* 2 S 



00 



vo 



o 

00 






oo 



SO 



o 

SO 

r- 

00 



(N 2 

2 OS 






S 



o 

00 



rn 

00 



00 CT\ 



i 
B 

>« TJ 

g| 



rn 



O 00 



j: os 



as 
so 



v> t- 

I I 



e 


n 


^ 


Ov 


Os 


©\ 



ON 



6 



1 

00 
SO 
CM 

m 



g 

ca 

a 
B 



T3 
<D 
00 
00 

o 



o 
H 






443 

Table 4-22. Vegetation types that burned during 1954-1955, and the types of vegetation 
that occurred in the burned areas by 1977. 



Vegetation Type that 




Vegetation Type 




Burned during 1954- 


Proportion* of 


Occurring in Burned 


Proportion of 


1955 


Sampled Area 


Areas by 1977 


Sampled Area 


Bare Ground-Urban 


0.4 


Upland Pine 


6.0 


Wet Forest 


40.4 


Needle-Leaved 
Evergreen 


0.3 


Open Water 


0.1 


Bay 


1.4 


Prairie 


13.7 


Cypress 


2.8 


Shrubs 


39.1 


Blackgum 


0.1 


Upland Pine 


6.3 


Bay-Cypress 


1.3 






Mixed-Cypress 


2.3 






Cypress-Shrub- 


7.1 






Prairie 








Mixed Pine 


0.2 






Herbaceous Prairie 


3.4 






Aquatic Prairie 


12.9 






Shrubs 


16.0 






Scrub 


7.6 






Scrub-Shrub 


9.8 






Shrub-Pine 


0.7 






Shrub-Cypress 


3.1 






Shrub-Bay 


6.1 






Shrub-Prairie 


17.0 






Scrub-Pine 


0.6 






Scrub-Prairie 


1.4 



The photo interpreted area included 58% of the refuge. 






444 



Table 4-23. Vegetation types that burned after 1955 and before 1990, and the types of 
vegetation that occurred in the burned areas in 1990. 



Vegetation Type that 

Burned After 1955 and 

Before 1990 


Proportion of 
Sampled Area 


Vegetation Type 

Occurring in Burned 

Areas in 1990 


Proportion of 
Sampled Area 


Upland Pine 


56.5 


Upland Pine 


0.3 


Needle-Leaved 
Evergreen (wetland) 


3.3 


Dense Pine 


30.2 


Cypress 


7.4 


Sparse Pine 


23.4 


Cypress-Shrub- 
Prairie 


1.6 


Clearcut-Sparse 
Pine 


0.02 


Aquatic Prairie 


0.3 


Pine-Cypress- 
Hardwoods 


16.2 


Shrubs 


8.8 


Mixed Upland- 
Wetland Shrubs 


0.03 


Scrub 


3.5 


Briar-Shrub 


0.1 


Scrub-Shrub 


2.4 


Mixed Wet Pine 


1.6 


Shrub-Prairie 


11.3 


Bay-Shrub 


9.7 


Scrub-Pine 


1.0 


Cypress-Gum- 
Shrub 


11.7 


Scrub-Prairie 


3.6 


Loblolly Bay 


0.6 






Shrub 


2.1 






Gum-Bay-Cypress- 
Shrub 


2.9 






Sedge-Fern- Water 
Lily 


1.1 



445 
affected by impounded water during high water conditions; this area decreases with 
declining water levels. The Cypress Creek watershed has also had elevated average 
water levels in 5140 ha since the sill was constructed, but during high water level 
conditions this area may actually drain more rapidly due to a reversal of the water surface 
gradient towards the Suwannee River (see Chapter 3). The following discussion 
addresses vegetation changes occurring in these areas. 

During 1952-1990 nearly all of the change that occurred in each vegetation type 
in the western and central Suwannee River Sill impact area (see chapter 3) was to wet 
forest (Table 4-24, Figure 4-8). Most of the wet forest in 1990 in this region was 
composed of gum-bay-cypress-shrub, cypress-gum-shrub, and loblolly bay (Table 4-25), 
and most of the change occurring during 1952-1990 was to these types from shrub-bay, 
shrub-prairie, shrub, and scrub, during 1977-1990 (Tables 4-26, 4-27). Conversion to 
shrub types in this area was primarily to bay-shrub and other shrub-wet forest 
associations. 

Prior to sill construction, vegetation change in the sill-affected area occurred in 
shrub, wet forest, and prairie vegetation types. Much of this transition can be attributed 
to succession following logging. In the sill impoundment-affected area during 1977- 
1990, changes in vegetation compositions were occurring at a much greater rate than 
those changes occurred during 1952-1977 (Table 3-11). Rates of changes outside of this 
area were also greater during 1977-1990 than 1952-1977 for wet forest, shrub, and 
prairie. However, nearly all upland pine change that occurred during 1952-1990 was 
complete by 1977. In the sill impoundment impact area wet forest area initially 



446 

Table 4-24. Vegetation changes occurring during 1952-1990 in the river floodplain area 
most likely affected by the sill's impoundment and in the Cypress Creek watershed area. 
Minimum mapping unit for the comparison is 240 m. Values are % of the vegetation 
class in 1952 occurring in the specified class in 1990. 



Area and Class 
in 1952 


Wetland 
Forest 
in 1990 


Shrubs 
in 1990 


Prairie 
in 1990 


Upland 

Pine 

in 1990 


Open 
Water 
in 1990 


Bare Ground- 
Urban 
in 1990 


Floodplain 
Area 














Wetland Forest 


77.3 


20.8 


1.2 


0.7 








Shrubs 


93.4 


4.2 


1.5 


0.7 








Prairie 


47.4 


33.2 


19.2 


0.3 








Upland Pine 


73.6 


9.9 


1.0 


15.1 





0.5 


Open Water 


99.5 


0.5 














Bare Ground- 
Urban 


100.0 

















Cypress Creek 
Area 














Wetland Forest 


57.0 


42.2 


0.1 


0.6 








Shrubs 


53.1 


42.0 


0.2 


4.7 








Prairie 


47.2 


52.3 


0.5 











Upland Pine 


12.3 








87.7 








Open Water 




















Bare Ground- 
Urban 


100.0 


















447 



c 




> 
c 
u 



00 

o 

I 

E 

t 
u 



u 



C 

o 

'5b 
u 

c 

•o 

c 

o" 

On 
Os 



ON 

oo 

2 
°c 

3 

-o 
oo 

c 



3 
5 
o 
o 
u 

00 

c 

g 

■S 

5 

c 
o 

-2 = 
o '55 

o £ 

B s 

J? 5 









00 « 

i i 

O 1) 
3-° 

•SP a 



448 



c .E 
u 

a.i 

£ 



£3 O 



^ t> « 
» so c« 

— o — " 

« a 1 

£ «g "S 
tLts s 

>> a 0< 

^^ £ 

J* g u 
5 fe 12 

few*-' 

rt a> l_. 

J3 r. § 

ts I *P 

O 3 .E 

E 9. & 

22 = e 
« « ^ 

loE 

£• c 'a 
f pi 

S eh: 

og.E 

St ?5 o- 
2D — 8f 

> a E 

2 E 8. 

P 3 

c 
■a to 

3 v) 
"O to 

3 u •& 

§ d O 

u b 9 

o ■ J 

8 "S ° 

C b ig 

III 

I?. a 

Um ■ — 

• o. t 
™ U 5 

-* s 8 

•5 T3 «n 

«J C On 



o 

c 

3 



■ 




















' c 




















s £ l 




















i- u "■ 




o 


o 


o 


o 


o 


© 


o 


e 


l f ■ 




















Q a- 




















■ 




















ulture- 

awn 

1990 




o 


o 


o 


o 


o 


o 


o 




z — = 




















M 




















< 




















■re 
iund- 
ban 
1990 




o 


o 


o 


o 


o 


u"> 


in 


o 


u - 5 — 














o 




d 


M 




















£ 




















2 © 




















to 2 




o 


1*1 


tn 


>T 


q 


— 


Ov 


(7» 


i. ~ 






fi 


d 


vd 


•— 


vd 


vo 


r-i 


5 S 




















% 




















£0 




















k 




















V 








































1 * © 




















Sedge 

ms-W 

Lily 

in 199 




o 


— 


o 




r<x 


d 


5 

T 






















V 




















m 




















<■! _1 




















Mix. 

Wetia 

Pirn 

in 19' 




o 


d 


o 


s 

d 


d 


^; 




d 


i 




















Gum-Ba 

Cypress 

Shrub 

in 199C 






p> 


»r> 


1 


°! 


f^ 


r- 
v^ 


in 




o 




a; 


*■■> 




vd 


vO 




>, 




















2g 




















b ON 




o 


CN 

d 


o 


d 


IN 

d 


d 


n-s 


in 


« B 




















it" 




















i 








































OB n c?v 




















s |S 




o 


Ov 


o 


u-t 


m 


wv 


O 
•A 


o 


E ffl e 






"" 




d 




ri 


»*i 




s 









































| E E 




o 


d 


o 


o 


o 


00 


O 


d 


ffff J 




















■o © 




















1 ■ £ 




o 


3 


o 


rs 


o 


© 


O 


o 


D .£ 






d 




d 


d 






d 


Vegetation 

Class and 

Area 

in 1952 


4: 
#1 


•Q 

B E 
U 3 « 

1 II 

BUB 


C -_ 
5 1 

i i 


k 

e o 
a " 
O ? 


I 


.2 

t« 


01 

c 

£ 

■a 
e 
n 

a 

s 


■ 
■ 

u «e 
1 e? 


"5 

ȣ5t? 



449 



1 

e 
I 



■ 

•8 
h 



C v 0\ 

- « = 

ye.— 



2 o I 

lis 

•c J ■ 



£ e « £ 



j St 



r 

09 



M* i" On 
IE 3 " 






fill 

So 1 *- 



k. ON 

* - 

■ S 



— e 

g >> On 

1*3 



I 



.11 

s. — 



I 



5 -a 



_2 I on 
S * £ 



S ■ a f 

1 " S 2 

■ ■ 5 •• 



o o o o o o 



o o o o o o 



o o o o o o 



O O L: o 2 



o o o o o 



o o o e o 



o o o o o o 



5 



E = .c 
« *- 1- 



1 1 



II 



t 

0. 



j J i 

C — S a X 

te 6 H < 8 



■ I 3 



■d 

1 



o 
u 

I 
I 



I 






:= S « S 

2 = S " 

o 2 «■ n 



i ill 

u X 



i 5 = 12 



ill 

5- £ « 



1 

c 


<U 


© 

I 


1 


EL 


£ 




i* 








? 




m 


_E 



450 



? 






i 



in 2 a 

6 2 8 



o 
d 



d d 



T 

d 



e 
d 



e 

o 



o 
d 



d 



2 S 

00 Z 



d 



fil 

3 >> c 



111 

flu 



II 

as 



if * 



3 I 2 



Sf - < "" 






O <N 






o r- <n 

m — vp 



e-e 

o a 



11 



e 



I! 

o £ 



1 



ij =e ff j m 



3 85 ■ ■ © 

* a s- •< 2 



t 









■ E u 

6fi ft 3 







c 


c 




















*1 

.o e 


■ 
E 


Jj 


5- 




o 


r-^ 


p 


d 










< u 














































c 
























■ 
























E 

■= 
c 

I 


SS 


r-< 


















■ 

e 
H 


5 

SB 

E 

4 


c 
c. 


o 
o 


o 




r- 
in 


(N 










1 






















■ 


2 


1 

* 


© 

1 


_ 


o 


© 


00 


p« 


VO 


c* 






c — 










d 


d 




d 




£ 


Ft 

U 




Jj 


















■ 

o B 


1 

c 


2 


0* 


o 


o 


o 


o 


o 


o 


o 




*§ 


c 

* 




a 


















1 

c 


4* 

c 






d 


o 


o 


- 


d 


T 


d 




m 

■ 
e 


■ 
c 


= 
■ 




d 


o 


o 




SO 

T 




!-^ 






e 


^ 




o 


o 


>n 


O 


o 


rs 


d 






B 


_C 






















X 






















j»> 


2 


1 




















2 


Eg 


m 




00 


o 


o 




V3 


»n 


»*i 




3 


i 

- 


c 




















■ 


2 


e 




















1 




£' 




r*i 




o 


>o 


VD 


<*■> 


^ , 




M 




SO 


o 


vo 


00 




O 


00 




E 


s 




IN 




T 


r4 


in 


•*T 


<N 




u 


3 

9 
























3 

E 


| 




<N 




00 


q 




fn 


m 






■ 




IN 


o 


d 


rj 


O 


oi 


(N 






b 


c 




■V 




»n 


■fl- 




r~ 


^r 






« 
























S3 
























k 


s 




















S 


■ 


s- 


















■o 


4) 

1 








O 


o 


o 


o 


o 


o 


o 


1 
























3 
























B 


U 


^ 


o 






















o 






















■ 


r> 


















C 

o 
o 

1 
1 


2 * 

II 


0> 




o 


o 


o 


e 


o 


o 


o 
























<n 


e 

1 

> 


■D 














B 


M 




■4 


C 
SB 


c 


r* 


•o 








£ 


n 


•/•ofTota 
Area in 
Class in 
1990 


n 


■ 

m 

g 


1 

< 


r, 




fc. 

B V 


•s 

e 

a. 


■ 
2 

-e 


B 
OS 

c. 


H -< S 



























451 



452 



c£ 
|.S E 

"i §3 

§ -£ S 
S ° <* 
E .2 g 

•—■SO 
c« S .22 

<u > E 
x: d o 

1 hS 

aj <zi 3 
>> % <u> 

m *a .E 

^ > S. 

w 01 n 

§ | E 

B % E 

Six. E 

.5 3 «5 
jfl O *5 

~ tfcj S 

2 o 5C 

O ON 

« T5 — 
js .52 "I 

+■• on Q. 
C C 1 

•5 o g 

*T E S 
D te 2 

II 1° 
flj 



3 
"O 



<u o 

III 

O *S *H 

» 5 5 

<u ? u 

SP.* 5- 

c 



03 



<D 



en 



-5 D D 

-> O ~ 

on S 

on "" 

B ex 



c 
o 



S 



VO -O -O 

<? S fi 

N M 3 

-Hog 

« «a c 

E-i <D 3 



B s Is 

B 



III 



Bl|{ 



« £ 



a 5 



5 



t 



■ 1 J 1 

ft* 2 M 2* 
•a t « — 

|2gj 



i £ K 

S « B 



1 5 -S I 

. C P ON 

E rI z 



1. o\ 









S 



S-s 



e» 1 on 

!•£ - 

3 5 



■2 £ 

■ 



" s e 
II; 



S s ° 



o 
d 



O0 

o 



d 



00 

_ o 



d 



(N PM 



e 
d 



o 
d 






00 

d 



00 

d 



■ 

S iff 

— ft* E «* 



3 

| 



t 

5 



£ 



PB U 



"g E 
I! 



£ 

4* 



E4J 

P- 2 « 



s 

II 

S a. 



■< CL. 



453 



9 

c 

1 

o 
o 

VO 



X) 

H 



3 I ■ £ 
E i 19 

6* .5 



B i — 
^1 



el if 

go J 



2 



■ S 

i £ 

•n 

ce 



ill* 



1 5 



JS" 5 



ce ■ 



? ■= 



I ft 



i i 



^ ? 

I; 

S B. >. * 

O I oa o 



Iff 



1 jl 



o 
d 



o oo 

— ci 



o 
d 



o 
ci 



e 



o 
d 



Ov oo 
**i — • 



o 
d 



00 

d 



00 

d 








VC 





o 
d 



o o o o 



o 

d 



1JS 



5 

ce 



x 



CO 40 






ce U 



op 

i 

e 

ce 



a ■ 

II 

ce a. 



B 

S 



« r _ 



£ 



f- = M 



2g 



1 e 



S ? < 



"i * a 



454 



1 

c 

I 

o 
o 

CN 
*T 

CO 



3 ■ 8 
urns 
E i IS 



lis 



i = 



3 5 



2 



X 

ea 



■ I kg 



III 





2 


0) 

* 


c 


>. 

CO 


E 




I 


F 


£- 




~* 


3 


5 


c£ 


c 



is 



i S-ST2 

UgCB . 



X E ON 



a * £ 



I lE 



o o o o 



o o o o 



oooo 



o o o o 



o o o o 



o o o 



oc 
d 



o o o 



6 6 



OOOO O O CO o o oo 



oooo © o oo o o oo 



oooo 



oooo 



g fig 

S Z J! H 



<S u 5 



E 1 1 Uj 

'£" Iff ^s£ 



I 

If 

S a. 



ft 

M 

< a. 



g 1 
O * 



2 
■ 



2 





















455 




ircut- 
arse 
ine 
1990 


o 


© 


o 


o 


o 


O 


o 


o 


o 


























h 






















iwn 
990 










































gric 
e-L« 
in 1 


o 


o 


o 


o 


o 


o 


o 


o 


o 




< 






















Bare 
Ground- 
Urban 
in 1990 


o 


o 


o 


O 


o 


o 


o 


o 


o 




■ 






















J 






















2o 






















f; 


o 


o 


d 


o 


d 


o 


o 




V! 

d 




■ .5 
•c 






















■ 






















£. 






















S?2g 






















Sedg 

Fern 

Water 

in 19 


d 


o 




o 


r» 


o 


© 




Ch 




■ 






















■a e © 






















ass 


*T 


\r\ 


r~ 


r* 


» 




rs 


r~ 








^6 






vS 




o 




\o 






l|3 








*T 


o 






T 


&■ 




Gum-Bay- 
Cypress- 

Shruh 
in 1990 


© 


© 


o 




e 


o 


r- 

d 




d 




£> 






















3§ 






















V — 


e 


o 


o 


O 


p 


o 


o 


00 


'T 




■*■* _ 


















d 




oi a 






















Z " 






















' i s 






















I a « +* 


o 


o 


© 


o 


o 


o 


o 


© 


o 




U S « = 
























o 


o 


o 


o 


o 


o 


e 


o 


o 




Jw J 




















t 






















3 
C 


C «■ ?■ 




















1 

O 


Upla 

Pin 

in 19 


o 


o 


o 


o 


o 


o 


o 


o 


o 


a 
























c 




4» 












is 




I 


.2 e 




s 




>, 




4) 




M .— 




TJ" 


1 ■ £ 




s 


■ 


1 
09 




c 




J 9 


II 

A _ 




Vegel 

Clas 

19 


11 

tr. -j-j 


JO 

2 

JS 
(A 


1 5 
ll 

09 U 


■ 

5 

■c 


ll 

(0 0. 


1 

2 

D 
OS 


1 4» 

II 


Total C 
Area (fa 
1990 


Area in 
Class in 
1990 















456 



1 

c 

1 

o 
o 

v£> 
<N 

■4 

(L> 

I 



— e r- 
M — wi r- 

H ■ g c\ 

J4«j 






£< 



i s 1 

I t E 

y 3 



a ^ to 



5 = 



5 



o 
I' 



J* I 



■ V> 









> 



o 
d 



o © 

SO — 



£ S S 
2 2 s 






vo „ 



rs 
O 



T _ 

o ~ 
ri O 



O 

d 



T o °? 



o 
d 



d — 



~ o 



o 
e 




111 

i % t 
z » u 



5 



Sf 



£1 



14 j 

u to a. 



a. 



« r e h ■§ 

11 li I 

< a. O ? S 






457 



c 
o 
o 

g 

— 

1 



£ ~ 2 b 
k I 1 x 



— . « — 

* e .s 



s fit 

u 5 



Sill 



»5 






%5$ 






8.1 



5 s 

It 



= s b 

« ;» 

> 



•J *o 






— o 



oc — 
d d 



d 



d d 



c 
d 



e 



NO 

d 



NO 

d 



o 
d 






p oo 
r-' \b 



\6 



d 



St 



o 



J * 

5 5 

Hi <n 



£ 

5 



II 



CO 

2 

5 



i e 

</) a. 



a. 
A 
5 



1 ■ 



u <£ 

15 * < 

1 | 



5 I si? 



j 

ll 



•a 



a. 



km = 

- ^ at 

r»; 

Z JS H 



458 



i 

1 

o 
o 

o 

i 

1 



H a I c\ 

;I 8 j 



■ it 

— ■ — 
2 £ e 



I 



c t I 

s M 

u ■ 



«ilj 

1 s = 1 

z a £ ■ 



SI 



is 



2 



g< 



o o o 



Z 2 r- 



w» in 



© '„ 



OC &v OC 



o o © 



o o o o 



E 






1 







>, 


2 
1 




«S 





II 



Ij 



2 I ^ 

8*5 = 



o o o 



o o o 



o o o o 



o o o 



b o o o 



e 
b 



o o o 



© © © © 



© © o o 



o © 



o o o 



*£; oc 

_ o 3 



— \C 00 

© — © 



o 



DO 

e 



U1 <N f*l 



s 



o © © 



o o o 



2! 
ft 



E 

3 
1 



I ft 111 



£ 2 



>. B. 
« >> 

ca u 



II 

Bo. < a. 



■ 4> 



a-* 



lie 



i « 




5 2 


2 


X <« 


£ 



459 



1 
1 

o 
o 

1 



H ■ g on 



IIP 

113 



1 1! 

s If 



•o i 1 a 

813 I 

2 s £ ■ 



1.1 






2 



1 - 

r 



A 4 



*e 






9 % 



o 2 



— « 



.2 (■ 
■ S * 



09 



X t - 

E I 2 

05 O (A 



f | 



I Ei iii 



H 1 ■ 

\ Is. 

* < tJ 



u 



460 



S .S 

vi 

Vj 

a 



E v> 
•3 £ 



c 

3 



8.5 



c 
o 

8 d 
o o 

> (N 



X! 


x: 


,s 


■+-» 


(4-1 


p 


>^ 


o 


r> 


Jo 




C/5 

'C 


<u 


to 


eO 

a. 


ifi 


CO 


E 




VI 

3 


o 
o 


>> 


CO 


r 


IU 


> 


•*-» 


.* 


. * 


u. 




V) 


n 


' ' 


to 


IS 


■*-» 






(/5 

o 


VI 
CO 


'E 


6 


o 


3 


CO 


•n 


DO 


'O 

trt 


| 

3 
O 


(- 

'a, 


c 


E 


CO 


& 


£ 



"3 O 

O 

O 
IS 



■I E 






c 
o 
o 

o 

c 



Os -Q — 

T D. C 

M OJ '" 

•oca. 
o c 
— r- 

f- v 

3 <U 
"3 -3 

C 

| "J 
3 2j JU 

O in * 

5 S <s 

cfl CO O 
1) % VI 

s ^ J 

en . rs 
(U o 



eo 

E 



a. 

3 
O 

feb 

c 

3 



C 

o 



2 II 

<u >> ST 

g>U « 

>j| 

*■• c 

S "3 ~ 

« 8 fi 

■S £ « 

g C3 on 









c 



5 i ■- 

*yl a. 



li 






t 






00 



lit 

I I & 



« £ 

3 



.2 I 



vo — 



e 

c 



00 

\6 



— N 

d d 



~ d 



OS 00 

— in 



wm r^ t 






d 



* _ 



00 



o 

00 



o 
d 



— oo 
d in 



d 



o 
So 



if 

w 3 



» 1 
I I 



_ ■ 

r 

1 

C 



£ 
£ 1 

* J? - 

»3 3 



g 1 

O & 



• t- 



Pji 



a g £ o ■ ■ p 



li 



It 

U 3 



461 



o 
o 

I 
I 

i 
«* 

I 



1 I 
11 


o 


o 


o 


o 


o 


o 


o 


h 

S = 
* 5 

II 


d 


d 


e 


o 


o 


T 


d 


c 

E 

1 

1 


0* 


o 


o 

d 


e 


o 


In 


r^ 


Cypresj- 

Shruh- 
Prairie 




o 


6 


o 


o 


§ 




I! 


o 


o 


o 


o 


o 


o 


e 


■ 
















o 


o 


o 


o 


e 


o 


o 


e 

3 
1 
















o 


o 


o 


o 


o 


o 


o 


















M 

■ 

E 

5 
















o 


o 


o 


o 


o 


o 


o 


i 


rs 


o 


d 


o 


o 


hi 


M 


Needle- 
Leaved 
Evergreen 


O 


o 


o 


o 


O 


e 


o 


h 

9 


q 


o 


Oj 


vi 


e 


<N 


m 


Vegetation 

Class in 

1952 


e _ 
Si m 


■ 

•c 

c 

a. 


C 

E 


■ 
e 

£ 

B 

— 

5 


B u 
ii 


■ = 

■ 

« 9 

1 IE 








462 



o 
o 

CN 

« 



« 












































— •» 












































m - 












































a in CI 

n!952 






o 


VD 
On 




T 
fi 


ON 
IS 

T 


v 




o 






















d 


$ t~ 












































< 












































■ 












































■ 












































•> rt 


















































o 


n 

00 












o 






















_ 


Fj 








oo 




n 


On 


c?> 






s 






















H 


















































o 






d 




d 




o 


NO 

no 






o\ 














o 


- s 


















































o 


o 




o 


o 


o 




o 


o 






o 














o 


1 *» 












































II 






c 


00 




HI 


d 


d 




o 


NO 

r- 






00 

r-' 














o 


(A Q. 












































i fe- 
ll 






o 


00 






no 
d 

r*1 






o 


00 

^- 

00 
f*1 




















o 


CO 












































*> 


















































o 






*T 


oc 
d 


o 




o 


IN 






»■■) 














o 


2 £ 

s - 






o 


rsl 

d 




O 


d 


r*i 




o 








d 














© 


<8 <« 






o 










d 




o 








On 














o 












































A 












































2 






o 


00 






NO 
r-' 


rn 




o 


5 
•<»■ 






NO 














© 


j 






















(S 






















2 

M 






o 






o> 


od 


no 




o 


00 

IN 




















o 


(A 












































II 






o 


O 




o 


o 


o 




o 


o 






O 














o 


c 


K 










































J c 
-** ~* n 


•3 

4 






■a 














t 






"3 
o 
















Vegeta 

Class 

195 


T 


1 


ill 


e 

is 


1 

o 


41 


n 

2 


■o 
c 
es 

a 


4* 
B 

c 




< 






e 


Si 


on 


Hi 


ill 

OB sis 



463 



1 

{ 

? 

t 

rr 

■a 



M 






















— ' 






















9 * 






















of Tot 
a in CI 
nl952 


**"i 




^ 


r~ 


_ 












* 




«d 


vd 


d 


o 










#|" 










































M 






















B 






















«■ n 
























O 




w^ 




„ 


o 










3*3 


"" 






M" 














H 






















42 






















£ m 


o 




o 


o 


o 


o 




o 


o 




j u 






















(A 0. 






















■S i! 






















5 5 


o 




o 


c 


o 


o 




o 


o 




1* 






















a 






d 

X 


9> 


o 


o 










•Si C 






















Is 


>»■ 






es 


o 








T 




TT 




o 




1 


o 




0^ 


t^ 




■ 
























o 




00 


U"l 








v> 


>n 




O 




d 


ml 


o 


o 




r- 


d 




•b >. 






















te (J 






















i , 






















5 5 


o 




o 


o 


o 


o 




o 


o 




a a, 






















■ 






















it 


Cvl 






vt 








00 


o 




K 




o 


00 


o 


o 




— * 


O* 




«•> V 








<s 








fl 


— « 




jC 






















E 


a- 




^ ^ 


~~ 








*n 


o 




*d 




00 


vd 


o 


o 




f«i 


rs 




fl 


*** 




(N 








v> 


<*■» 




■ 






















« 






















£ 


o 




o 


o 


o 


o 




o 


O 




f 






















■ 






















a e> 


o 




o 


o 


o 


o 




o 


o 




o* 






















B 
















■ 






S B 
= - M 


V 














E 


*o3 




13 a 


a 


1 


■ 

£ 

A. 


E 

»3 


B 

.2 « 

s £ 


s 
u 
a 



1 


< B 


5 I 


■ 
52 



464 
increased during 1952-1977 and then remained nearly constant during the next 13 years, 
whereas shrub, prairie, and upland pine areas were nearly halved during 1952-1990 
(Table 3-10). Prairie was replaced with wet forest and shrub. Upland pine increased 
during 1952-1977 and then decreased. These rates of change inside the sill 
impoundment impact area were less than those observed in the swamp overall during 
1952-1977, and greater than those in the swamp overall during 1977-1990 (Table 3-1 1). 

During 1952-1990 in the sill-affected Cypress Creek watershed area, most 
vegetation change was to wet forest, and more than half of the loss in prairie area was 
due to replacement by shrub (Table 4-24). Most of the shrub in 1990 in this area was 
composed of loblolly bay-shrub, and cypress-gum-shrub and mixed wetland pine 
(primarily slash pine and pond pine) made up the wet forest type (Table 4-25). The 
change to these types during 1952-1990 was primarily from scrub, scrub-shrub, shrub- 
bay, shrub-cypress, shrub-prairie, and cypress-shrub-prairie during 1977-1990 (Tables 4- 
26, 4-27). Prairie conversion in this area was primarily to bay-shrub and cypress-gum- 
shrub associations. 

At least half of the changes occurring in the Cypress Creek watershed area 
occurred during 1952-1977. In contrast, most vegetation change in the remainder of the 
swamp occurred during 1977-1990, although most upland pine conversion occurred 
during 1952-1977 (Table 3-11). In the Cypress Creek area, prairie and upland pine 
coverage declined during 1952-1977 while shrub coverage increased. Wet forest 
declined by half during 1952-1990, and shrub coverage continued to grow during 1977- 
1990 (Table 3-10). 



465 
Discussion 

The Okefenokee Swamp landscape has been affected by disturbance episodes of 
various types, intensities, and extents during the past 150 years. The responses to these 
disturbances have varied temporally and spatially. The overall structure of the landscape 
at the beginning and end of this 150 year period was relatively similar. Total proportions 
of the swamp area in wet forest, shrub, and upland forest associations have not changed, 
nor have the general locations of these communities in the landscape. Within this period 
shrub communities have been replaced by wet forest, prairie by shrubland and wet forest, 
and wet forest by shrubland and prairie. However, these changes are on a shorter 
temporal scale than the overall structural persistence of the system over the past 150 
years. There has been some alteration in the species' compositions of these structural 
types, however. Although there are many areas that have returned to their pre-logging 
composition, there are other forested regions of the swamp where cypress and shrub- 
prairie were probably more abundant prior to logging and loblolly bay, loblolly bay- 
shrub, and blackgum-loblolly bay coverages have increased since logging occurred 
(Table 4-28). Some of this change resulted from the early 20th century logging. 
However, evidence suggests that disruption of the natural fire regime may also be driving 
this landscape evolution. 

Cypress and pine were the predominant species logged from the swamp (Izlar 
1984). Their return to the landscape has depended on the presence of a seed source, 



466 



8 W) 

2 .S 

■5 s 

42 ao 
6 o 

> a. 

E -= 

5*1 

* J 

S • r 

a £ 

8 2 

■3 a c 
£ 'Z .2 

si j 

2 a a 

c (S > 

'C *f '-S 

3 ai 

73 S I 

E 3 £ 
M § o 

J gj 

o 3 C 
c c « 
||g 
13 c § 

o s i 

•So E 
_c 2 cH 

111 

t/i 

a> tn 

So 2 ■ 

a "o & 

c 8 * 

o <u 



o 



1.9 

en 

in 



BO 



u — 



ed 



3 

E 

CO 

c 
o 



o ,2 

"O tn 

(N f*3 O 

1 O C 






u 



4g 5T P 



' en 



"5 ■ 5 

a i •£ 2 



llll 



lis 

T J .8 



o 
d 



e 
d 



o 
d 



e 

o 



D 



d 



2 






•c 
■ <n 



Kt* .2 e\ 
l§3i 



llll 



llll 



L. ON 



c 
d 



o 

d 



o 
d 



o 
d 



O 
d 



£- 




















Gum- 
Maple-Bays 
in 1990 


o 










OS 

d 




so 


(« 


Ogeechee- 
Cypress 
in 1990 


o 




IN 

o 

d 






d 




o 


o 

d 


Upland 

Pine 

in 1990 


o 
d 




o 

d 






d 




d 


rr 

d 


Prelogging 

Vegetation 

Class 

(1855-1890) 


a 

a 
O 


M 
>> 

a 

ffi 

■ 

c 
« 


■ 

D ■ 

| s 

5c? 


a 

2 
in 


■ 

11 
Si 


1 

si 


■o 
e 
■ 



CN 

d 



o 
d 



M 

£ a. 



X 




«c te 



1 
U <n 



5 



467 



4 I If 
ffJJ 



■o 4 1 r, I 



a ' 



51 



52 



— o 

o 



e 



o 






■g as 
3 M 



X 

£ (A R 

3 



If 

CO 



111 



< O -S 



u 



w = e 



O 
o 



6 






e 
d 



a 
d 



o 
d 



■ 
as 

e 2- 

o s 



Ef 



5 



u 




o 

o 



* 



e 



o 
d 



4 i 



« — 

o a 



8 e 
II 



i! 



IS 



E 

f 



468 
either from coppice growth (cypress) or water (cypress) or wind-dispersed (pine) seeds. 
Cypress seeds, like many woody wetland species, do not survive extended periods in the 
wetland seed bank (Demaree 1932), although they may survive submergence for up to a 
year (Applequist 1959). Since water flow is limited in much of the swamp to stream 
beds, most of the cypress regeneration distant to stream and river floodplains, where 
seeds are water dispersed, has probably occurred by coppice growth and local coppice 
production of seeds. Where these sources have been eliminated, cypress have also 
disappeared. Even where seeds have been available, the conditions for germination and 
seedling survival may have been limiting. Pond cypress's requirement of abundant light 
precludes it from areas already populated by shade-producing shrubs and trees 
(Terwilliger and Ewel 1986, Best et al. 1984, Hamilton 1984, 1982), and the seed's brief 
survival when submerged (3-12 months) eliminates it from establishing in areas with 
long hydroperiods following seed rain (Applequist 1959, Demaree 1932). Wetland pine 
(P. serotina and P. elliottii) was logged from sites with shorter hydroperiods and lower 
water depths, and replaced by bays, blackgum, and shrubs, to some extent a result of fire 
exclusion by humans. Again, seed dispersal and survival may have been a limiting factor 
as pine seed trees were removed, replaced by shade-producing shrubs, blackgum, and 
bays, and extended flooding occurred with impoundment near the sill (Pritchett 1979, 
Shriver and Fortson 1979). Once these competitors have become densely established, the 
possibility of cypress and pine reoccurrence depends on the additional factor of a severe 
fire which, as is discussed below, has not occurred in the swamp since logging occurred. 
Cypress and pine establishment did occur in some areas during and immediately 



469 
following logging. Terwilliger and Ewel (1986) and Ewel et al. (1989) reported highest 
densities of young pond cypress during the first few years following logging in North 
central Florida cypress domes. They also reported recovery of composition in logged 
domes within 45 years of logging; most of these domes had been selectively logged, and 
all showed evidence of recent burning. Neither of these factors are true for most of the 
Okefenokee Swamp forested areas. 

Between the pre-logging period (1850-1890) and 1952 there was an increase in 
prairie, shrub, and upland pine communities and a decrease in wet forest coverage. This 
change was mostly due to revegetation by shrubs in sites that were forested prior to 
logging. Although individual shrub species can not be identified in the 1952 
photographs, it is likely that by 1952 most of this shrub community was dominated by 
titi, with fetterbush, Virginia willow, and soapbush as secondary components; these 
species are colonizers of recently exposed peat and require short hydroperiods, low water 
depths, and high levels of light (Hamilton 1984, 1982, Deuver 1982, 1979, Cypert 1973, 
1972, 1961). Other common shrub species in the swamp, such as hurrahbush and 
climbing fetterbush (Pierus phillyreifolia), are more shade tolerant and dominant in the 
forest understory, and probably did not occur in abundance in the 1952 communities 
where overstory growth was sparse. Loblolly bay, blackgum, and cypress seedlings may 
have already become established by 1952 in titi communities where sources of 
regeneration were available. Their presence could not be confirmed from the type and 
scale of photographs available for 1952, but subsequent remote sensing data from 1977 
and 1990 indicate that these species were present in some areas occupied by shrubs in 



470 
1952, although not detected in 1952 as mature trees. By 1990 hurrahbush and climbing 
fetterbush were found in satellite image classification ground-truth plots (see Chapter 2) 
that contained gum-bay forest. 

Between 1952 and 1977 nearly all of the swamp was burned by wildfires, and by 
1977 most burned areas were replaced by shrubs, shrub-prairie, scrub-shrub, or wet 
forest. Proportions of prairie and upland pine remained constant during this interval; wet 
forest eliminated by fire was replaced by shrubs, primarily shrub-bay. In areas that were 
previously logged, replacement was equally by wet forest and shrubs by 1977, and prairie 
and upland pine to a much lower extent. The variety of wet forest and shrub types was 
much greater by 1977 than recorded during the pre-logging period. Although this might 
be an artifact of the pre-logging survey notes, resolution, and map, it probably also 
indicates the effect of the logging on the landscape composition. Large areas of 
relatively continuous vegetation types (probably densely canopied areas with shrub 
understory) were dissected by logging tramlines. In some area cypress remains only 
outside the scars, indicating that it was probably out of reach of the logging equipment. 
Revegetation in the logged portions to a different composition has probably increased the 
species complexity in those areas today. Logging introduced patches and edges where 
none previously existed, and created an edge type (that of a break in the forest canopy 
due to large-scale removal of trees) that previously had not occurred in the swamp, or 
occurred only after severe burning. 

In most cases the wet forest association in a location during 1977 contained 
species also found in that area before it was logged; differences occurred primarily in 



471 
species dominance. This suggests that the source of regeneration was present after 
logging, but competitive interactions, site changes, and altered disturbance regimes may 
have affected the species composition in 1990. Thus, removal of the dominant wet forest 
species (primarily pond cypress, with some pond and slash pine in perimeter areas) by 
logging and modification of factors maintaining the system, such as the fire regime, have 
resulted in replacement by another species; where sufficient seed source remained and 
light was abundant, such as outside the logged area or at the logged fringe, the 
community probably more closely reflects the composition of the pre-logging era. 

By 1990 the most common vegetation types in the swamp were cypress-gum- 
shrub, bay-shrub, and gum-bay-cypress-shrub, evolving from the shrub, scrub, and shrub- 
scrub dominated landscape of 1977 (Table 4-29). These associations occurred prior to 
logging, although cypress was dominate instead of bay and blackgum. The period 1977- 
1990 was notable for the numerous wildfires, which were quickly extinguished, and 
therefore were limited temporally, spatially, and in intensity. In the absence of severe 
fires, swamp communities have followed the successional sequences proposed by 
Hamilton (1982), and species requiring severe fire for maintenance are being replaced 
with those that thrive when fire is eliminated from the system (Figure 4-9). The fires that 
have occurred during the past 150 years have been litter-reducing, but have not been 
severe enough to cause long-term (i.e., century) changes in the swamp landscape 
structure, such as changing forests or shrubland to prairies or lakes. During the past 
decade of wildfire management, even these litter-reducing fires have been suppressed. 
This is permitting a fuel accumulation that could support an extensive, severe fire during 



472 



I 

E 

o 
,c 

E 

o o 

to m 

60 

> | 

J3 B 

2 a 

o S 
nP to 

* <S 

§5 

■ c 

■i-s 

a 5. 
E 

ea 

8 £ 

on £ 

El 

on _d 

00 2 
c 
■£ o 

3 OS 
T3 On 

a,"" 1 

E fc» 

5 -° 

8L 



TO 

> 



CO 

.* o 
2 '-5 

OX o 

<L> 00 

-^ <u 

. w 
oo£ 

■g I 

o *-> 

OJ +5 
2P 60 
|.fi 

x: oo 
o c 

g a 

.2 o 

. ON 

On *— i 

^ J 

t -1 c« 



3 t u J 
S « B 



2 o 

u 
< 



£bSJ 

«Jl3 



1 ? m 
j J" 



Mil 



I I ■ I 



<P I ? s 



i & 



2£ 

ij 



i I 


>. OS 

« — 

■ M 


1 

JB 

u 


5 — 

E 0\ 

6 — 
•-■ c 
U - 


s 

OB 

c 
3 


e 

£ i 



.2 I- 
i It 



d vS 



O 



— On 



3 2 









0C 

NC 



e 



'■i oo *^ vi 



o 
d 



oc 



o 
d 



IN On 






o 
d 



c 
d 



O 



a. § 

g lip 

3 Z ± u 



E 

& 



f 



£ § 



B el 



I MB is m I 



2 X « X 

= O. < CL 



4) *- 



A 

2 



2 



473 



1 

e 

1 

o 

o 

I 
I 

OS 

<? 

1 



6*- 



IIS 

■t - = 



t 5 I £ 

1 1*5 

o 3 a 



ca m .5 



3t ? .2! o\ 
f tl" 

<8 E J .S 



S£ £ = 



5 x g 

2S2 



u 



£•<« 



is 

41 -« 



I g- if S 

U S «c 



S p.- 



S 3 



2 S ft 

> 



o 
d 



00 W") 

— 6 



o g 



a 









<N — 






O 

d 



o 

d 



d 



o 
d 



o 
d 





e 

£ 


« 


£ 




I 

(V 




ii 


^ 


i I 


£l 




A 


i V 


^ 


E 


is 


t 


£s 


e 


1 


<A «! 


on 


IT: U 


XA 


ir. a- 


<n a. 



474 



L I I r 
c I * 2 

w a 



1 1 1 5 2 

S a J* J o 



| 

Eg 



8 - 



52 

to J 



I <0 "" 



£ 

E 
= 




1(3 



-= on 

co 2 
as 



ill 



5 e 



e 

= «- 

« 5 f~ 

2 2 c\ 

B — ~« 









O T <N 

— do 



oo \o 






O 

6 



NO 



nc 
oc 



c 
o 



ON 

d 



c 

d 



o 
d 



e 

c 



NO 

d 



<s o 

d — 



o 
no 



vO ^ — 



"! — f> 

S - H 



o 






NO 

c 



ON On 

00 — 



aq v. 
iH no 

CnI — 






c 

c 



a 
- 






1 I 



5- & j 

ee u oa cc 



I E 



II! 



s 

If 



< Q. 



I 



i 5 



5 



475 



Hi 



lillf 

S J J IS 

* B ^ »5 -S 



- ? 

1 5 



v £ 



£ 



d 



d 






>o — i 



s .2 



1 M - 

3 j 



U 5 



II 



s 



o -- 






oo r- 

— oo 



d 



e 






d 



o 



> 



05 (O 



i 
5 

JB 



03 U 



CO 

5 



ef 



e 

£ 

(Jo u5 CL 



476 



m 







00 

(3 


o 








o\ 






§8 


on 




CD 


o 


c 




fc 


2£ 


S3 




CD 




r- o t-» 




•** 


U, 


r- on f- 


22 


s 


n> 


On On Q\ 


•^ 


1) 


u 


^■■H ^"* *»* 


U- 

o 


13 

O 


0) 

> 


II II II 


2 2 

ii ii 


00 
II 


D00 




m 



3 U oi 2 

2 .2 g E 

© co ■ £ 

& * 3 O 



8." 



1 i 

-O X) 

T3 

^ <D 
S3 60 
^ § 

^3 x 

II 

§ x 
1 ^ 

CD *j 

ft w 
§ I 

of 

€ E 

c £ 

•- o 

DO O 

c 



c 
c 

o § 



to 

co 

ID 
O 

o 



0,45 

ll 

S -s 

3 S 

CO C, 



o 
Q 



CO 

I 



On C 

tU 03 
u. *■* 
-j (U 

Bo 60 



T3 <D 

O 2 

° o 

3 E 

w 

si 

E 8 
3 X 
co e 

<d .E 

s S 

» t 

u o 
* .E 

i E 

c « 

"™ CO 

co C3 

1 S 

■e o. 

ll 

"O co 

8 E 
| §> 

i s 

< s 

& x 

ON • -] 

co e 

o 3 

•s 3 

l| 

/-s « 

PJ5 

K CO 

O co 

c i 

3°.S 
« S3 

S * 

.8 f £* 

« gg 
a. o o 
£>.E w 



477 
the next extended drought. The coupled effects of historic logging with this altered fuel 
load and fire regime are currently creating a landscape with similar structural 
appearance, but different species composition than the system present prior to logging. 
The predictability of the system's response to future wildfires will decrease as the 
landscape continues to evolve, driven by the altered disturbance regime and species 
composition. It is likely that a severe, extensive fire will burn during the next 50-100 
years (Yin 1993) and alter much of the swamp landscape. 

The Suwannee River Sill has extended hydroperiods and increased flooding 
depths in approximately 15% of the swamp (Chapter 3), primarily east and northeast of 
the sill. The Cypress Creek watershed (approximately 3% of the swamp) has also 
experienced an increase in water surface elevations, although it is unclear whether this is 
attributed directly to the sill or to other changes in the watershed drainage since sill 
construction. The sill's spatial effects are most extensive during high water periods; 
during low and average water depth periods water is impounded in a small area around 
the sill, and within the primary drainages. Most of the decrease in forest cover and 
increase in shrub and prairie cover in the sill impact area during 1952-1977 probably 
occurred during the wildfires of 1954-1955, and during 1960 when the marketable timber 
was removed from the area anticipated to be impounded immediately east of the sill. In 
the Cypress Creek watershed changes during 1952-1977 were from prairie and shrub to 
shrub-prairie and shrub-scrub types, following the 1954-1955 fires which burned 
primarily in prairie and shrub types. Expansion of forest and decline in shrub coverage 
during 1977-1990 occurred while the river floodplain area was almost continuously 



478 
impounded. There were occasional intervals during this period when the ground surface 
was intermittently exposed due to low precipitation or sill gate structural failure, 
permitting germination of cypress, blackgum, and bay seeds. The same trends were 
observed in the Cypress Creek watershed, where forest-shrub mixes replaced shrub- 
dominated communities. Stands of cypress and blackgum that established since the sill 
was constructed currently occur throughout the floodplain area impounded by the sill. 
Seedlings that developed into these stands probably survived flooding because they were 
dormant when it initially occurred, or they had attained sufficient height to tolerate 
extended flooding once it occurred. During 1993 and 1994, dry periods in the 
impoundment area extended long enough to allow establishment of cypress and 
blackgum seedlings that survived late season flooding. Although woody shrub species are 
present in the seed bank and might also germinate during these drawdowns, seed survival 
is probably low (see Chapter 7). Other than buttonbush (Cephalanthus occidental is), 
those that do survive to germinate can not tolerate the long hydroperiods and deep 
flooding in the area east of the sill. Therefore, shrubs rooted in the ground probably have 
gradually been displaced from inundated sites in the river floodplain sill impoundment 
area, and only those on exposed stump surfaces, floating logs, and pond cypress 
buttresses have survived. Outside the floodplain region of the sill's impact, wet forest 
area is also gradually increasing while shrub coverage decreases. Cypress Creek 
watershed vegetation is following this trend. This displacement is occurring more 
rapidly inside the sill impact zone than outside, possibly because the extended 
hydroperiod and extreme water depths limiting germination of shrub species more 



479 
frequently occur, and established trees are tolerating these conditions. Recruitment of 
trees is sporadic, however, and probably occurs only during extreme, extended 
drawdowns. Throughout the swamp, vegetation succession is occurring in the absence of 
fire, away from fire maintained associations to those that flourish without fire; however, 
the driving functions differ. Hydrological manipulation is the controlling function in the 
sill impact area, whereas fire suppression is directing vegetation dynamics elsewhere. 

Okefenokee Swamp vegetation changes occurring during the past 150 years 
suggests multiple scale processes that differentially affect the landscape structure and 
composition. The temporal and spatial effects of seasonal storms on swamp vegetation 
are small; severe wind creates localized disruptions in vegetation communities, with little 
effect on the landscape. Hurricanes have frequently passed over or near the swamp; 
unlike hurricane effects reported by Putz and Sharitz (1991) in bottomland hardwood 
forests in coastal South Carolina, accounts of widespread damage to Okefenokee Swamp 
vegetation have been infrequent during the past 150 years (C. Trowell, pers. comm.). 
Increasing abundance of bay species in the forested regions may increase the swamp's 
susceptibility to hurricane damage, however. Putz and Sharitz (1991) found greater 
damage to bottomland hardwood species rooted in elevated sites, and concluded that leaf 
size and permanency which creates wind resistence may also have made these species 
more prone to wind damage. Drought appears to have a greater potential effect than 
hurricane damage on Okefenokee Swamp vegetation composition and distributions. 
There appears to be a cycle of severe drought that probably occurs every few hundred 
years and lasts for several years (Yin 1993; Chapter 5 this volume). During these 



480 
infrequent but extreme droughts, wildfires are ignited and may burn across the entire 
swamp and cause temporally and spatially extensive changes in the swamp landscape 
composition (Cohen et al. 1984, Cohen 1975, 1974, 1973a, 1973b). Evidence in the peat 
suggests this occurred every few hundred years throughout the swamp since initial peat 
accumulation 6,500 years ago (Hermann et al. 1989, Cohen et al. 1984, Cohen 1975, 
1974, 1973a, 1973b). Where these fires are sufficiently severe, new lakes and prairies 
might be created as trees and shrubs are killed and peat is burned. This is the 
hypothesized origin of several lakes and prairies present today (Cypert 1961, Hopkins 
1947). Where they are less severe, communities undergo less extensive change, and 
recovery to pre-fire conditions probably occurs in a few years to decades, as has occurred 
in the Cypress Creek watershed. 

Intermittent with these semi-century drought and fire cycles are less severe, less 
extensive fires that may not burn throughout the swamp (Yin 1993), and may be less 
spatially uniform in their severity. This results in fuel reduction in some areas and short- 
term structural or compositional changes to the landscape in other areas, and maintains 
the moving mosaic of communities in different developmental stages, similar in total 
composition to that present before burning. The fires that burned through the swamp 
during 1954-1955 and 1931-1932 were probably this type (see Hamilton 1982). In the 
absence of disturbance an area might change from prairie to forest within a century, with 
a regular series of succeeding species and associations determined by general site 
hydrological conditions, light availability, and propagule source (see Chapters 6, 7). 
When disturbance types that reduce or eliminate system variability are introduced, such 



481 
as logging, fire suppression, and impoundment, the system may become less resilient to 
successive disturbance events, and the type and scale of the system's responses may 
become unpredictable; a new stability domain may emerge as the system reorganizes 
(Holling 1995, 1986). It is possible that Okefenokee Swamp is currently responding to 
these types of disruptions with replacement of cypress-dominated forests with forests 
dominated by loblolly bay and blackgum. 

Logging, which ceased 60-100 years ago, affected the swamp forest structure by 
removing trees and the seed source (primarily pond cypress in the swamp interior and 
pines in the uplands and swamp perimeter) and creating large areas suitable for shrub 
growth. Logged areas have frequently burned, and these fires may have been more 
severe where logging debris had accumulated; this may also have affected the subsequent 
species composition. Although areas where pond cypress was logged have since 
reforested, the composition has changed from pre-logging so that fire frequency and 
behavior and hydrologic processes (e.g., water flow, hydroperiod, evapotranspiration 
rates, etc.) have also probably changed. Recent alterations in the swamp fire regime due 
to wildfire management and use of controlled burns are probably also modifying the 
swamp landscape structure and composition, by suppressing the "mosaic maintaining" 
smaller fires as well as those that might burn throughout the swamp due to drought 
conditions. 

Changes in the swamp landscape from prairie and shrubland to forest, and from 
cypress forest to bay-gum dominated forest, will remain unchecked if wildfire 
suppression continues, unless a large, hot, uncontrolled and uncontrollable fire occurs. 



482 
Shade tolerant, fire intolerant, and fire suppressing species are replacing cypress, which 
requires fire to remain in the landscape (Hamilton 1984, 1982, Best et al. 1984), may 
promote spread of fire by its features (e.g., shaggy bark, serotinous leaves and cones), 
and can survive fires that do not burn into the peat (Ewel and Mitsch 1978). This large- 
scale alteration in species composition in the landscape due to extensive early 20th 
century logging and disruption of the driving functions of fire and hydrology, may be 
reorganizing the swamp to a new stability domain. As the spatial variability of the 
landscape that gave the system resilience to extensive perturbations such as fire and 
drought disappears, an Okefenokee Swamp landscape will emerge that differs from that 
developing historically. 



CHAPTER 5 
FIRE IN OKEFENOKEE SWAMP 



Introduction 



Wildfire is integral to creating and maintaining certain vegetation communities. 
Effects of fire are over many spatial and temporal scales, with local to landscape-level 
responses occurring instantly and possibly continuing for years. Peat based wetlands are 
but one of several ecosystems that depend on periodic fires to shape the composition and 
structure of the landscape. Human-induced changes in community and landscape 
composition and structure potentially alter fire intensity and periodicity and affect the 
predictability of response to fire in these systems. Persistence of an ecosystem that 
evolved with and is naturally maintained by fire depends on maintenance of its natural 
fire regime. Without the fire disturbance, species are displaced with those that are better 
adapted to the changing conditions, and the successional sequence proceeds. If fire 
suppression is successful, a fire-dependent system ultimately can change into a system 
dependent on the absence of fire, and the intended protection actually leads to the 
system's demise. Environmental adaptations determine the suite of species that can exist 
in a landscape; their evolution to tolerate the existing and changing conditions and 
disturbances determines when a species appears and disappears from the landscape. 

483 



484 
Although the culmination in a climax community rarely occurs across the landscape 
because of natural, restructuring disturbance events such as fire and drought, there is the 
potential to alter the composition of the interim communities and affect the predictability 
of the system's response, such as with artificial disruptions of hydrologic regimes or 
species composition due to logging. This disruption not only modifies the species 
composition but also can affect the behavior of the system in future disturbance events. 

Evidence of fires in the Okefenokee Swamp exists from the oldest peat 
accumulated in the swamp interior (Cohen et al. 1984, Cohen 1975, 1973a, 1973b). 
Many of these fires probably originated in the swamp as lightning strikes and were 
eventually extinguished by saturated peat or precipitation. Their origin may also have 
been as lightning strikes in the perimeter longleaf pine {Pinus palustris), slash pine (P. 
elliottii), and wiregrass (Aristida spp.) communities, which are dependent on fire for 
establishment and maintenance (Hermann et al. 1989). With suitable weather conditions 
these fires could spread into the swamp matrix where they might be extinguished by 
saturated peat, or continue to bum into the swamp interior. Fire frequency varies 
spatially in the swamp. Throughout the swamp there is evidence of a large fire 6,000- 
10,000+ years ago, and intensive fires in 13 periods or roughly every few hundred years 
since then (Hermann et al. 1989, Cohen et al. 1984). Some areas, however, have no 
charcoal bands in the peat, suggesting they have escaped severe fire (Hermann et al. 
1989, Cohen et al. 1984). Residents of the swamp area noted extensive drought and fires 
during 1838-1840, 1844, 1860, 1910, 1932, and 1954-1955 (Trowell 1987, Izlar 1984, 
Hopkins 1947). Many of these fires occurred during the spring months of March-May 



485 
during a drought following a severe freeze (Trowell 1987) and were extinguished by 
precipitation. Inhabitants of the perimeter uplands have frequently used fire to improve 
forage quality for livestock during the past few centuries. Indians occupying the area 
during the past 10,000 years were also known to use fire in land management (Trowell 
1987, 1984a), although they had no ability to suppress fires ignited during droughts 
(Hermann et al. 1989). 

Fire within the swamp has created landscape structure by removing areas of 
forest to form aquatic prairies, maintaining herbaceous and shrub prairies, and removing 
accumulated peat to create aquatic prairies and relatively deep depressions where lakes 
form (Cypert 1973, 1961). The last extensive fires responsible for substantial prairie 
initiation were probably in the mid- 1800s (Cypert 1973, 1961, C. Trowell, pers. comm.), 
although small areas of prairie (each <200 ha) developed in the 1954-1955 burns 
(Hamilton 1984, 1982, Cypert 1973, 1961). The result throughout the swamp's history 
has been a moving mosaic in the landscape of vegetation communities, which become 
less dependent on maintenance by fire as they are removed from the effects of fire. 

The usual sequence of swamp vegetation succession with peat accumulation is 
from aquatic prairie dominated by fragrant water lily (Nymphaea odorata) to herbaceous 
prairie of yellow-eyed grass (Xyris spp.), broomsedge (Andropogon virginicum), and 
sedges {Carex spp.), and in the absence of severe fire or extensive inundation, to titi 
(Cyrilla racemiflora) and eventually pond cypress (Taxodium ascendens) and swamp 
blackgum (Nyssa sylvatica v. biflora) forest (Hamilton 1984, 1982). The early 
successional species do not readily germinate and establish in shaded conditions, which 



486 
they create as they mature. Depending on the species, various hydrologic conditions are 
also required for establishment to occur (see Chapters 6 and 7). Fire occurrence and 
effect is decreased by extended hydroperiods; some areas of the swamp have always been 
aquatic and show no evidence of fire (Cohen 1973b). Inundation has always inhibited 
establishment of flood-intolerant species in these areas, except on elevated tree islands. 
If absence of severe fire is prolonged, and effects of inundation are reduced by 
organic soil accumulation, the forest composition changes to shade tolerant species, such 
as fetterbush and hurrahbush, and cypress is accompanied by loblolly and sweet bay, and 
dahoon holly (Best et al. 1984, Hamilton 1984, 1982). As conditions become more 
shallow and shaded, cypress might be dominated by these hardwoods, which are less fire- 
tolerant and unlike cypress, have features that reduce fire susceptibility, such as 
sclerophorous leaves and smooth bark. Eventually only a severe, hot fire or mechanical 
removal of trees will permit invasion by earlier succession species (Hamilton 1984, 
1982). Throughout the swamp the patchy distribution of vegetation types and evidence 
of past fire in the peat suggest that historically, fires were of variable size and intensity; 
their effects varied spatially due to the hydrologic environment and existing vegetation 
composition, and a shifting mosaic of communities has been maintained for at least 
6,500 years (Cohen 1975, 1974, 1973a, Cohen et al. 1984). 

Fires burned 1 14,935 ha of refuge and 60,705 ha of surrounding commercial and 
state land during 1954-1955 (Chapter 742, Public Law 81-810, 70 Statute 668). Although 
some of these incendiary and lightning-caused fires originated in the swamp and spread 
to the surrounding upland, many of the ignitions were on perimeter land, as landowners 



487 
set backfires to control the spread of fire away from the swamp and into the surrounding 
uplands (S.M. Reeves, pers. comm.). The severe drought conditions at the time enabled 
the upland and swamp fires to spread, burn into the peat, and remain active until 
precipitation extinguished the flames in June 1955. It was believed that had the 
Suwannee River been impounded at the time of these fires, the effects of the drought on 
the swamp would not have been as severe, the peat would have remained flooded, and 
the saturated peat would have extinguished the fires in the swamp before they spread into 
the perimeter. Additionally, protection from fire was viewed at the time as integral to the 
swamp's integrity. The sill, therefore, was intended to "protect the natural features and 
very substantial public values represented in the Okefenokee National Wildlife Refuge, 
Georgia, from disastrous fires and protect against damage from drought" (Chapter 742, 
Public Law 81-810, 70 Statute 668). 

In addition to questions of the effects of the sill on the swamp hydrology 
addressed in Chapter 3 are questions of its effects on the swamp's natural fire regime. 
The sill is one of several fire control methods used by refuge staff. A perimeter road and 
fire break are maintained to arrest lateral fire movement. Active suppression of lightning 
strikes and incendiary fires, and a winter controlled burning program are also part of the 
refuge's fire management plan. Each of these management activities alters the natural 
fire regime by affecting fire periodicity, intensity, and spatial extent. These affects might 
be direct, as by impounding water and active fire suppression, or indirect by altering 
species compositions and therefore modifying fuel loads, moisture regimes, and fire 
susceptibility, or affecting the landscape structure with canoe trails and roads and thus 



488 
altering fire movement in the landscape. The actual contribution of the sill to its 
intended objectives must be discerned before deciding the sill's fate. 

This chapter examines fire sizes, distributions, and frequencies during the period 
of available records, 1855-1993, and discusses effects of the current wildfire 
management activities on the swamp landscape. My objective was to determine whether 
changes in swamp vegetation distributions identified in Chapter 4 can be attributed 
directly to wildfire suppression or reduction caused by impoundment effects of the sill 
structure. 

Methods 

Wildfires and Prescribed Fires 

The Okefenokee National Wildlife Refuge has maintained a wildfire occurrence 
log since the refuge was established in 1937. Prior to that date, fires were noted in 
regional newspapers, diaries of the area's inhabitants, and records of logging companies 
active in the swamp and perimeter (Trowell 1987). The multiple sources have resulted in 
records of varying content and quality. Information such as date, time, location, ignition 
source, fire size, and fire distribution maps are included in records of some fires; others 
note only general descriptions. Fire data used in this analysis included as many of these 
sources as possible. No doubt some fires were not recorded, or the information was 
overlooked while compiling data for this effort. The various sources also contribute a 
range of error. However, the spatial coverage database created from the available 



489 
records are believed to be generally representative of the overall wildfire history during 
1855-1993, so that trends in fire occurrences and vegetation responses can be examined. 

Descriptions of wildfires occurring prior to 1937 were taken from maps and 
documentation of Trowell ( 1987). He gathered most of this information from 
newspaper accounts and logging company records of fires in the area. The original, hand 
drawn maps in this documentation were not spatially referenced. The spatially 
referenced maps created for this analysis were screen digitized from these maps over the 
background of the registered, rectified, and merged 1990 SPOT satellite image. The 
tramline map Trowell 1994, 1983) (see Chapter 4) was also used to reference these fire 
locations. Digitized polygons of fire coverages were adjusted until the size was within 
10 ha of the general burned areas estimated in Trowell (1987), or until the polygon 
appeared to approximate that in Trowell (1987) where sizes were not reported. 
Trowell's maps (Trowell 1987) were intended to show general location information of 
the fires and not actual acreage, so there may be error in these estimates due to uneven or 
incomplete burn. Therefore, they estimate maximum estimated burned area. Fire 
information from 1937-1993 was retrieved from refuge fire and biological reports and 
annual narratives. If fire maps were available, these were transposed onto 1966, 7.5" 
USGS 1:24,000 topographic quad maps and digitized to create wildfire coverages. Each 
fire year became a separate coverage, and if fire polygons overlapped in a single year, the 
coverage was divided to remove this overlap. 

Some (22%) of the wildfire records, representing 2% of the estimated total area 
burned by wildfires during 1855-1993, included locations and size estimates but no 



490 
maps. Sizes of these fires ranged from 0. 1 to 101 1.8 ha. These fires were plotted as 
points using the location information for approximate placement, and then the points 
were buffered (surrounded) to create a circle with a radius that would result in the 
estimated burn area. This provided general fire location coverages for the wildfires 
without using actual reference maps. Although vegetation and topography were not 
considered in this approach, the method was believed to be acceptable given the small 
total acreage involved (x = 105. 1 ha, range=0.01-101 1.8 ha). Buffered points were saved 
as wildfire-year coverages unless wildfires in a single year overlapped; these were saved 
in multiple coverages for the year to spatially isolate each wildfire. Fire records (22) 
providing no size estimate or location were not mapped. Ignition source and date were 
recorded for all wildfires where available. All fires caused by humans, whether 
accidental or resulting from an escaped prescribed burn, were recorded as incendiary, 
since the distinction was not complete for all records; other causes of fire were lightning 
strikes or unknown. 

Formal prescribed burn records were not kept by the refuge staff until 1973. Prior 
to this date there are anecdotal accounts of prescribed burns in the refuge records. A 
map and description were included if available, as for wildfires, for all prescribed burn 
records prior to 1973. The burn compartment, block, and unit designation were used for 
locating all prescribed burns following 1972. A compartment-block-unit designation was 
assigned to each prescribed burning polygon transposed from refuge hand-drawn maps to 
7.5" 1:24,000 USGS quads. The dates of prescribed burns were added as coverage items, 



491 
so that the prescribed bum coverage contained all individual fires identified by burn year 
and month. Prescribed burn summaries are calculated from this coverage. 

All spatial comparisons among wildfire coverages were made using ARCINFO 
(version 7.0, ERSI, Inc., Redlands, CA 92373) and IMAGINE (version 8.2, ERDAS, Inc., 
Atlanta, GA 30329) GIS analysis software. Fire polygons were converted to grids of 
10x10 m cells using ARCINFO-GRID (ESRI 1992), and imported into IMAGINE 
(ERDAS 1995) for generating reports of each comparison among burn periods. 
Graphical comparisons in fire size and frequency were made to identify temporal trends. 
Wildfire Occurrences and Vegetation Types 

Distributions and compositions of swamp vegetation prior to logging around the 
turn of the century, and during 1952, 1977, and 1990 were examined in Chapter 4. Maps 
developed for the vegetation change analyses in Chapter 4 were compared with maps of 
areas burned during 1855-1993 to determine vegetation types that probably occurred 
where fires burned, to estimate the types of vegetation that resulted in the burned areas, 
and to determine if areas where vegetation change occurred were also where wildfires 
occurred. Fire maps were also compared with the tramline map developed in Chapter 4, 
to ascertain the association of wildfires with previously logged areas. Fuel loads at the 
time of each of the vegetation maps (1855, 1952, 1977, 1990) were also estimated using 
the fire behavior models of Anderson (1982) to compare available standing fuel (not 
peat) with fire sizes and numbers. Four models were used to estimate fuel amounts for 
different vegetation types prior to each of these periods. Time since last fire is not 



492 
considered in these models; the models provide estimates of burnable fuel relative to 
each vegetation type, but do not necessarily indicate how much will actually burn 
(Anderson 1982). The models can be used with weather and site condition information 
to make general predictions of fire severity. Fuel model values used for swamp fuel load 
calculations are listed in Table 5-1. 

All spatial comparisons among wildfire and vegetation coverages were made 
using ARCINFO and IMAGINE GIS analysis software. Gridded fire polygons were 
imported into IMAGINE for generating reports of each comparison among burn periods 
and vegetation maps. Fire maps were similarly compared with the tramline map. Fuel 
load calculations were made with vegetation type-area calculations from the vegetation 
maps developed in Chapter 4. 

Results 

Fire Sizes. Frequencies, and Causes 

Ehiring 1855-1993 approximately 96% of the swamp was burned by 249 
wildfires. Approximately 302,079 ha burned during this period, much of it repeatedly, 
and 161,583 ha burned since the swamp became a National Wildlife Refuge in 1937. 
The largest fire was 1 15, 020 ha, which occurred during 1954-1955. This fire was 
actually a combination of several separate ignitions that merged into a common burned 
area. The average area bumed by individual wildfires was 1017 ha; 81.5% of the fires 



493 



C4 

00 
ON 



c 
o 

VI 

i- 

OJ 

c 

< 



to 

C 

_o 

3 
_o 

u 

1 

o 



,3 



-o 

u 

3 



<u 

T3 

O 

E 

3 



— 
X) 
ea 

H 



T3 










03 










O 










al Fuel L 
(kg/ha) 


00 


"1 


in 


en 


H 

* 


NO 

00 

o 


o 


00 




00 
NO 


00 


•w 










© 










H 










■o 










4> XI 
00 ■** ,^*. 


NO 


m 


no 


NO 


% v S> 
= Q 


© 


00 


© 


o 


U. 










^3 v <a> 




tj- 






s a« 




00 




t-- 


■*■ .2 =5 


o 


o 


o 


NO 


£ '"§ 9 




fN 




ON 


.£ £ £ 




^~ 




00 


-9 




*— * 






"U 










S3 










© 










- u « 


© 


od 
o 

CM 


</1 


oo 


ad Fue 

<0.64 

(kg/h 


NO 


NO 

m 

m 


ir! 

NO 


■ 










a 










_» s 










- e£ 










o 


r- 


o 


cs 


igii 


vi 

N© 


^ 


© 


00 


© ^ T3 w 


On 




ON 

o 


- es es 




(N 


*"* 


*™ H 


© B 










- a 




















o 










m 


t 


vO 


r^ 


1 










V 










a 




■+-* 


e — 




,>> 




n 


5 « 
Z i 




H 

a 
# © 

40 

2 


a 

'os 
E 


o 

EXh 

M 

H 


*-* 

£ 

o 


• 

DC 
> 




1 

a. 
>> 


c/3 U 


u 

c 



B 
6 

E 

CO 

'■E 

a. 



1— 

C3 

ew 

C 
O 

C/3 

c 



3 
— 



494 
burned less than 1% (1600 ha) of the swamp (Table 5-2). Prior to 1937 most wildfires 
were reported in March- July (Figure 5-1), and the largest fires began during April, June, 
and July (Figure 5-2). Other fires (excluding the large fires of 1931-1932) also burned in 
June and July (Figure 5-3). Following refuge establishment, the peak in bum area was in 
March, largely due to the 1954-1955 fires (Figure 5-4), and peak fire numbers occurred 
in June-August (Figure 5-5). If other fires (excluding the 1954-1955 fires) are considered 
after 1937, the peak in burn area appears to have shifted to July-November, with the most 
fires reported in July-September (Figure 5-6). 

Lightning was the ignition source of 122 fires during 1855-1993, and 73 fires 
were known to be incendiary; 54 fires were of unknown origin. Since 1937 there have 
been 222 wildfires recorded in the refuge; 70 of these were known to be of incendiary 
origin, 1 1 1 were ignited by lightning strikes, and the sources of 41 fires are unknown 
(Figure 5-7). Lightning-caused fires occurred in March, May, and July, and incendiary 
fires in April during 1855-1937. Following refuge establishment in 1937 lightning fires 
were reported primarily during June-September, and incendiary fires occurred during all 
months, particularly January- June (Figure 5-7). 

Frequency and seasonality of wildfires have not been uniform across the decades 
since refuge establishment. During 1938-1959 wildfires were reported during all 
months, and were most frequent in January and March-June. Excluding the fires of 
1954-1955, the larger fires were in March-April and August-November. Most of these 
fires originated as lightning strikes (Figure 5-8). During 1960-1979 wildfires were 
reported in all but January, and were evenly distributed among months. Ignition source 



495 



Table 5-2. Summary of wildfires in the Okefenokee Swamp National Wildlife Refuge 
area, 1855-1993. 





1855-1993 


1855-1959 


1960-1993 


Number of wildfires 


249 


98 


151 


Mean burn area 


1379 


3723 


159 


(ha) 








Variance of Burn 


l.lxlO 7 


2.9xl0 7 


1.1x10 s 


Area (ha) 








Maximum burn 


114935 


114935 


8407 


area (ha) 








Minimum burn area 


0.04 


0.1 


0.04 


(ha) 








Mode of burn areas 


0.04 


4.05 


0.04 


(ha) 








Number of lightning 


122 


33 


89 


ignitions 








Number of 


73 


30 


43 


incendiary ignitions 








Number of 


54 


35 


19 


unknown source 








ignitions 








Number of fires 








burning portions of 








swamp: 








<1% 


203 


62 


141 


1-4% 


11 


9 


2 


5-9% 


2 


1 


1 


<10% 


216 


72 


144 


10-24% 


1 


1 





25-49% 


1 


1 





50-74% 


1 


1 






Table 5-2--continued. 



496 





1855-1993 


1855-1959 


1960-1993 


75-95% 











Number of fires 


30 


23 


7 


with no area 








measurement 








Number of 


77 


51 


26 


unmapped fires 








Mean area of 


1737 


135 


64 


unmapped fires (ha) 












497 



0) 


(0 


3 


CO 


c m O 


$ .5 c 

O "D C 


C C *j 


£ 0) £ 


c O O) 


D £ D 






"~ 


p 






__ 


u 



- o. 



'e. 



-*o< 



'N 



-'o. 



-*» 



©c 



-<?„ 




r «* 



- v <?, 



* 



f, n „ , 



0* 



5 
c 
o 



*- 



so 

o 

v-> 

oo 

00 

a 

"C 

3 

« 

b> 
a 

<u 

00 

a 

(2 

,4) 



CO 

c 
_o 

03 

2 

<L> 

<u 

o 

c 

u 
M 

o 

25 



C 

a 

K 

en 

T3 



m 






<o 



l 

CM 



paiJodaysajnp||MP# 



o 
u 



CO 

u 



I 









498 




- O, 



- ft 



o 
o 



o 

00 



o 

CD 



O 



o 

CN 



spuBsnom 

S9J!iPI!M Aq paiung (eq) eajy 



o, 



«o, 



'A/ 



-** 



H 



o< 



%, 









V 



so 
tn 

Ov 

i 

»r> 
v> 

oo 

a 

'B 

3 
T3 
to 
1 



p 

— 
S3 

u 

00 

a 

ft. 



c 
o 

1 

Z 

O 

c 

o 

1) 



T3 

<u 

E 

3 
X) 
ea 

I 

o 
H 



0> 

ha 






499 










- o* 



"o. 



->o. 



->% 



-<?„ 



""?, 



- o, 




--** 



© 

CO 



CM 



o 

CM 



in 



r 
o 



«*L 



^ 



9* 



"^ 



%, 



in 



spuesnom 
S9J!^p|j/\A Aq paujng (et|) eajy |ejoi 



ON 



ON 



ID 



I 
I 

00 

c 



5 | 
o I 



NO 

tn 

On 

i 
>/"> 

oo 

00 

c 



o 

IS 
T3 



-a 
E 

3 
-C 

CO 

aj 
E 

o 

H 

I 

— 



500 



■ 

(0 

3 
CO 

tig 

lis 

COO) 

■D £ 3 



a 



I 



o 



o 

CM 



O 
O 



O 
00 



O 
CD 



O 



O 
CM 



(spuesnom) 
S8 -"iPHM Aq paiung (eq) eajy |B}oi 




— — - 

: 

t 



*>, 



*jfc 



%, 



^ 



"^ 



^ 

s 

K 



£ 

c 
o 



'«>, 



ON 

o 

bo 

c 



•a 



X) 
V 



1 

o 

s 

2 

I 

(4-1 

<D 

O 
i> 

c 



= 

1 

1 



501 



on 
On 
On 




- o. 



iQ 



'<V 



*>. 




^ 



* 



9 «* 



K 



V 



c 
o 



i 



,1> 



ca 

Z 

D 

o 

C 

-^ 
O 
<u 



I 

g 

"O 



— 
o 



O 

in 



o 



o 

CO 



o 

CM 



p8Viodaysaj|ip|!M# 






■ 



U* 



502 



- o, 



<fc 



■■•■'■ ' ■- '■■■■'■^■lI ....■:.•:..■ ■. ■■.. ..:.:.■. . :., : 



■■■,..■■.■■ 



I'** 



\-!o. 




(spuesnoqjj 
saj^pn/vy Aq paujng (eq) eajy |ejoi 



ON 

I 

IT) 



U 

o 

£ 

DO 
S 

•6 

5 

x 

m 

ON 

60 

c 

1 

2 



X) 



I 

I 



u. 



503 



i 


■ 


3 


(0 ffi 


known C 

endiary 

htning 

wn Caus 

iary 

ng 


c o a o o c 

3 £ j c £ a 


<B ■ • 1 2 .2> 


as o> a 3 JE _i 


"fc t: *§ ©" e af 


2 2 P a © o> 


re- 
re- 
re- 
efu 
efu 
efu 


Q. Q. Q. K E IT 


DDI 


...;,' 


... ,! 


1 




1 — [ 



V"" ' - ' 



..,,..,,]. *_ 



^a. 



'* 



*>, 




lllllllllliur — ~ 



II ' ! I ! I I ! I II ' . . ! ! ' ; ! I . , " ' ' ' ! I ' 1 1 y> 



- ! , ■* 




HI 



■461 



% 



9s 



%, 



<«. 



o 



o 



o 
co 



o 

CM 



pavioday sajwpuM # 



,i> 



c 
,o 

■4-» 

CS 

o 

c 

(4-1 

O 

m 
o 

■ 



00 

c 



•a 

s 

On 

i 

>r> 

oo 



00 

c 
o 






(U 
co 
3 
a] 
u 

9 

C3 

u. 

<U 

X) 

£ 



c 



2S 
. « 

CO 

D 






(2 



504 



• 

CO 

3 
I 

c ■ a 
j .2 c 

o "o "E 

coo 
3 = J 



DII 



H 



v 



** 



<% 



«■/-. 



»« 



»/ 



L 



I 






^ 



*to 



<& 



*A 



^ 



ff A 



*l 



*«J 



<^ 



% 



6"/ 



*fe 



*l 



§ 



o o o o o o o 

(N O 00 CD M" CN 



(spuesnoiii) 

saJJiPHM *<\ paujng (eq) eajy 



(0 

> 



%, .2 



i 

00 



00 

c 



en 

1 
I 



O 
<D 
.9 
5 

eg 

-^ 

o 

.c 

oo 
c 






s 

1 

•8 



I 

00 



- 



505 
was predominantly lightning (59%) (Figure 5-9). Since 1980 wildfires occurred during 
all months at 2-10 times the frequency of early periods, and were most frequent during 
April and June-August. During 1980-1989 fire sizes were not as small as they had been 
during the 2 previous decades (Figure 5-10). By 1990-1993 wildfire size increased 2-5 
times that of earlier decades. Ignition source for most wildfires during 1980-93 was 
lightning strikes (Figure 5-9). 

The prescribed burning program intensified in the early 1970's. Although much 
of the upland areas were periodically burned prior to this date, the record is inconsistent 
until 1974. Since 1974 prescribed burning has occurred annually primarily in December- 
February on large islands and perimeter uplands (Figure 5-11). The prescribed burning 
season usually corresponds to periods of high water levels (Figure 5-12). Approximately 
51,606 ha have been prescribed burned at various frequencies since 1953. 
Vegetation Chanpes Where Fires Occurred 

Most of the swamp burned during 1855-1952 (91%), and again in 1954-1955 
(83%). Approximately 13% of the swamp has burned since 1955. Areas that burned in 
the swamp prior to 1952 were primarily in cypress-gum-shrub (shrub is a mix dominated 
by titi, hurrahbush, Lyonia lucida, fetterbush, Leucothoe racemosa, soapbush, Clethra 
alnifolia, and Virginia willow, Itea virginica) (38.9%), greenbriar (Smilax spp.)-shrub- 
prairie (primarily sedges, broomsedge, water lily, and chain fern, Woodwardia virginica) 
(34.6%), and gum-bay (loblolly bay, Gordonia lasianthus, sweet bay, Magnolia 
virginiana, swamp red bay, Persea palustris)-cyvxess-shn\b (14.5%) (Table 4-18). 



506 




1 
■ 

3 

a 
° t 

tag 

o "5 1 

J* I .c 
c o O) 

3 £ j 

nil 




-e, 



% 



°e t 



*/ 



fc 



*a 



^ 



[ r^ 





»/ 



*•«, 






* 



*i. 



% 




«"/ 



*A 



<s* 



*« 



«* 



% 



<?/ 



*e, 



*/ 



o 

CD 



O 



O 



o 

CO 



S9J !1PI!AA jo Jeqmnfg 



m 

OS 
OS 

i 

IT) 
00 

60 

c 
c 

3 
T3 

a. 
E 

to 
I 

o 
c 
ft> 
u 

O 



do 

c 



i 



o 



s 

1 
I 



OS 

i 



- 



507 




m 

OV 

i 
<n 

00 
00 

c 

— 

§ 

a. 



J 

o 

c 



(spuesnoqjj 
SQJtfPHM Aq paiung (ei|) eajy |ejoi 



U 



I 



■ 



- 



508 




Prescribed Burning Compartments 

Sill 

Refuge Boundary 




5 5 10 Kilometers 



Figure 5-11. Prescribed burning compartments in the Okefenokee National Wildlife 
Refuge. 



509 



I 



88jy pequosejd Aq poujng (eq) eejy |e}oj_ A|q)uoyy 



o 


o 


O 


o 


o 


o 


o 


o 


o 


o 


§ 


■ 


o 


■ 


o 


14 


CO 


CM 


CM 



o 

s 



o 
o 
o 



__ I 

5Z • %• 

Kr."._. J» 




(tu) mdea jbjbm eBejeAv A|>|8eM|g 



U 

CO 

eo 



§ 



o 

"5 

c 

CO 

U 

<U 
<L> 

C 

c 

s 

3 

co 

T3 

c 

CS 

CL 

u, 
U 

CO 



eo 

Oh 

S 

I 

*-» 

CO 
La 
G 

* o 



U 



■S.2. 

G 

*» 60 
CO C 

a I 

!> wT 

> n 

s B 

■z J 

<D CO) 

« .9 



-o 
D 

'C 
5 

DO U 

tt, EL, 



fN 



in 
u 



510 
Revegetation in the burned areas by 1952 was primarily to wet forest (primarily slash 

pine, Pinus elliottii, pond pine, Pinus serotina, pond cypress, swamp blackgum, loblolly 
bay, sweetbay, and dahoon holly, Ilex cassine) (40%), shrub (41%), and prairie (14%) 
(Table 4-18). 

The fires that burned in 1954-55 were the first widespread fires to affect the 
swamp following creation of the National Wildlife Refuge. Wet forest (38%), shrub 
(37%), and prairie (19%) communities were burned, and were replaced with shrub- 
prairie (17%), shrub (16%), prairie (16%), scrub-shrub (10%), scrub (8%), and a mixture 
of wet forest (8%) and shrub-forest (18%) types by 1977 (Table 4-22). By 1990 these 
areas had become bay-shrub (23%), cypress-gum-shrub (28%), sedges-ferns-water lilies 
(10%), shrub (10%), gum-bay-cypress-shrub (9%), and loblolly bay (6%) (Table 5-3). 
During the 36 years following the 1954-1955 fires, burning occurred primarily in the 
non- wetland parts of the swamp. Upland pine (primarily slash pine and longleaf pine, 
Pinus palustr is) (57%), shrub-prairie (11%), shrub (9%), and cypress (7%) communities 
were burned, and by 1990 revegetated with upland pine (54%), pine-cypress-hardwoods 
(16%), cypress-gum-shrub (12%), and bay-shrub (10%) (Table 4-23). During 1990-1993 
approximately 13% of the swamp burned. Most of the burned area contained 
communities of cypress-gum-shrub (34%), bay-shrub (27%), shrub (9%), and upland pine 
(10%) (Table 5-4). An intensive effort was made to extinguish wildfires during 1990- 
1993. Most of the area burned in 1990 and 1993 was by lightning ignition when water 
levels were low. The previous period of severe burning (1954-1955) also occurred 
during low water conditions, and the initial ignition source of those fires was lightning; 



511 



Table 5-3. Vegetation in 1990 in areas that burned during 1954-1955. 





Percent of the Area that Burned in 


Vegetation Type in 1990 


1954-1955 that is the Specified 




Vegetation Type in 1990 


Upland Pine (includes dense and 


5.0 


sparse pine types) 




Ogeechee-Cypress 


0.01 


Gum-Maple-Bays 


0.5 


Water Lily 
Gum-Bay-Cypress-Shrub 


1.5 
8.9 


Mixed Wet Pine 


1.9 


Sedges-Ferns-Water Lily 


9.5 


Briar-Shrub 


4.1 


Agriculture-Lawn 





Bare-Urban 


0.1 


Clearcut-Sparse Pine 


0.01 


Aquatic Grasses 


0.04 


Open Water 


0.05 


Bay-Shrub 


22.7 


Cypress-Gum-Shrub 


27.8 


Loblolly Bay 


6.1 


Shrubs 


9.6 


Mixed Upland- Wetland Shrubs 


0.5 


Pine-Cypress-Hardwoods 


1.7 



512 



Table 5-4. Vegetation that burned during 1990-1993. 



Vegetation Type in 1990 


Proportion of the Area that Burned 
during 1990-1993 


Upland Pine (includes sparse and 
dense pine) 


10.0 


Gum-Maple-Bays 


0.2 


Water Lily 


0.1 


Gum-Bay-Cypress-Shrub 


4.8 


Mixed Wet Pine 


2.4 


Sedges-Fern- Water Lily 


5.6 


Briar-Shrub 


0.6 


Bare-Urban 


0.1 


Clearcut-Sparse Pine 


0.01 


Aquatic Grasses 


0.02 


Bay-Shrub 


27.2 


Cypress-Gum-Shrub 


33.9 


Loblolly Bay 


1.3 


Shrub 


9.1 


Mixed Upland- Wetland Shrubs 


0.1 


Pine-Cypress-Hardwoods 


4.6 



513 
the first fire of the period was extinguished, but later incendiary and lightning-caused 
fires burned in spite of suppression efforts. 

Fuel loads were estimated for swamp vegetation compositions in 1890 
(representing the pre-logging period), 1952, 1977, and 1990. Peat is not included in 
these calculations, which consider only standing vegetation. Fuel volumes using the 
Anderson models (Anderson 1982) indicated a gradual increase since 1952, following a 
decrease between 1890 and 1952. A severe, intense fire followed 2 of these periods 
(1954 and 1990), and several fires occurred during 1855 to 1942, when commercial 
logging ceased. There were approximately 7.5xl0 8 kg of burnable fuels in the swamp in 
1890, prior to logging activities. During 1855-1951 there were 65 wildfires reported. By 
1952 the fuel volume had increased to 7.9xl0 8 kg, which decreased following the 1954- 
1955 fires to 4.6xl0 8 kg of burnable fuels in 1977, possibly as a result of the 1954-1955 
fires. During 1952-1977 there were 54 wildfires that burned 137,170 ha. In 1990 the 
fuel load had increased again to 6.6xl0 8 kg, and from 1977-1993, 121 wildfires burned 
21,549 ha. The composition of these fuels by vegetation type are estimated in Table 4- 
20. Fuel loads of prairies and upland pine communities were fairly constant among 
1890, 1952, 1977, and 1990. Shrub fuel load was highest in 1977 and lowest in 1990. 
Wet forest fuel loads were highest in 1855 and 1990, and lowest in 1977. Wet forest area 
was composed of 55% cypress in 1855, and wet forest accounted for 79% of the fuel 
load. By 1977 cypress-dominated forest covered only 9% of the swamp area, but 
represented 23% of the total fuel load. In 1990 the cypress fuel load had increased to 
60% of the total, and cypress covered 32% of the swamp area. Much of the cypress 






514 
forest in 1855 was replaced by non-cypress wet forest by 1990, although wet forest total 
fuel load was similar to that in 1 855, when cypress dominated the wet forest area. 
Vegetation Changes Regardless of Fire Occurrence 

Changes in swamp vegetation were not dependent on the occurrence of fire. 
Overall proportions of changes in vegetation composition occurring during 1952-1977 
were generally similar between areas burned and not burned during 1855-1952, 1952- 
1977, and 1977-1990 (Table 5-5). That is, the occurrence of vegetation changes during 
1952-1977 was not necessarily dependent on fire occurring during 1855-1952, 1952- 
1977, or 1977-1990; similar transitions were indicated in the absence of fire as well as 
following burning (discussed in the previous section). Persistence of wet forest and 
shrub in 1952 and 1977 was not determined by burn history; however, compositions of 
these structural types differed between these periods. Most of the vegetation change that 
occurred in burned areas was from wet forest to shrub, or shrub to wet forest; types of 
changes in non burned areas were more varied (Table 5-5). 

During 1977-1990 vegetation changes where burning had occurred in 1952-1955 
were primarily shrub to cypress-blackgum-shrub or shrub to blackgum-bay-cypress-shrub 
(Table 5-6). This type of change also occurred in non burned regions, and persistent (i.e., 
never replaced by another vegetation type) wet forest was more common where burning 
had not occurred. Most of the fires in 1977-1990 were in areas of persistent upland pine. 
Where fires did not occur during 1977-1990, the primary type of vegetation change was 
from shrub to blackgum-bay-cypress-shrub . During 1990-1993 fire consumed areas that 






I 

ON 



On 

i 

>r> 

00 
60 

c 



</-> 




516 



H 
t I £ 

i- « i^ 

Sir 

5? IS 

CO 



z- = £ 

OS B X 



*A 



E2 



M 



es -r 



x 



r- 

"8 2 

2 2 

3 

ea 



3 
C 



51 E 

S-§2 

< -s § 

1 |2 



M 

CO. .£ «S 

1-52 

<S -o u") 

«; F» 



Z 'c H 

? -5 2 

<_ a 81 

e E« 



o 
o 



00 

cs 



CM 



SO 



O 

rn 



SO 



SO 



ON 

es 



SO 
© 



so 


O 


** 


CN 


00 


r^ 


ci 


<N 


© 


(N 




(N 


<N 


so 


O 


<N 



on 
© 



so 



r<> 



O 






«N 



O 
© 



© 



</"> 


t-; 


*t 


Ti- 


o 


-3- 


«o 


On 


O 


(*■» 


^ 


en 


<* 


CM 


o 


O 



so 



00 



o 

00 



o 



«n 

sb 



•O 









B ^ 



i 

■s 






> s 

§ o 
22 

£1 *l 
•B R 

cc _ £ 



e 

** 4< 

«- a. 

u. e 

- « 

£ 1 



.2 <r> 

XL t 

0> o 

a. u. 



| | 



e 
- r 



« 2 



O 

u. 



-r 

a 
>-. 
U 



S 2 






2 ■ 
O en «j 






e « 

J ;c 

E i 

(X. £ 

* E 

i£ bo 



B 
J <u 



S 

e 



es 
B 

O 



517 



T3 

I 

1 

O 

o 

I 
<A 

1 



M 






























rea 

urin 

990 






























<2t 

«- "C r- 


O 




© 


o 


O 


o 


O 




o 


O 




o 




O 


e it r~ 






























* I a 






























oa 






























+£ M 






























e e 






























% of Area N 

Burned durii 

1977-1990 


© 




o 


o 
o 


O 


o 


O 




o 


© 




o 




© 


M 






























. e 






























\rea 
duri 
1977 




































fN 












m 








^ 


% of. 
1952- 


O 




o 


O 

© 


O 


© 


O 




© 


O 
© 




o 




© 






























■ 






























*■ et 






























o c 






























% of Area N 

Burned durii 

1952-1977 


o 




o 


© 


O 


o 


O 




o 


o 

© 




o 




© 


M 






























Krea 

durin 

1952 




















c-> 








■* 


*"■ "S l/. 


o 




o 


o 


o 


o 


O 




o 


o 




o 




o 


° i i^) 








o 












© 








o 


S? E £ 

3 
BQ 


























































rr M 






























J I 
Z 'C M 

■ 12 






























% of Are 

Burned d 

1855- 1' 




















m 








<N 


o 




o 


© 


o 


o 


O 




o 


O 

© 




o 




©' 


■ 






























ffc 


o 
I. 


a. 


S 

u 


© 

u 


s 


a n 


r s 




e 
*• it 


e 
— 




o 




J 

CL 


b ej 

.2 £ 


I 


e 


it — 

9 E 

* J 

II 


i 
eg 


■ -5 


1 1 


11 




it .- 
*- cs 








3 
o 


*S ON 

M 


c 
I 


"B 
§ 

a 


O CO 


O co 


u 

C CO 


i e 


C 

'5 


11 


e 

a. 


E 

I 

a. 


c 

-*- 

1 


- 


5 1 


> H 


O 2 


oa 


£ 


O co 


o 


£ 


* 


a. cl 



518 



T3 

<D 

3 
C 

c 
o 
o 

I 
I 



CO 

H 



M 
























rea 

urin 

990 
























<: •« — 

«5 TJ t^ 


o 




O 


o 


o 


o 


o 


o 


o 


o 


o 


our- 
























S? E2 
























■ 3 
























■ 
























•r M 
























3 c 
























Z r 2 
























of Area 
irned du 
1977-19! 


~ 




00 


o 


© 


d 


d 


m 


o 


rn 

cn 


o 
d 


J?,g 
























M 
























. e 
























% of Area 
imed duri 
1952-1977 
























O 




tt' 


co 

o 


CN 

© 


es 


o 
d 


CN 


o 


d 


























OS 
























r w 
























© c 
























rea IN 

duri 
-1977 


m 




Os 


CN 


rf 


^o 


CN 


o 


CN 

o 


rt 


es 


%ofA 

Burned 

1952 


© 




<* 


© 


© 


d 


d 


VO 


d 


d 


CN 


M 
























% of Area 
Burned durin 
1855-1952 


o 






© 


o 


© 


m 
o 
d 


p 

00 


o 
d 


00 

d 


d 
























« w 
























.5 c 
























Z *c es 
























% of Are 

Burned d 

1855-1' 


r^ 




Os 


m 


<N 


t^ 


CN 


o 


o 


CN 


•-" 


o 




*— i 


© 


© 




d 


K 




d 


■ 












l 












■ 












■ 












5 b 

E H 


— 






£ 

05 


i 

£ 


CO 

**> "C 


4 
£ 

D 


g 

2 

- 


e 
I 

O 


1 

a 
3 


G 


I !C 


e 




e 


e 


o 


e - 


e 


e 


c 


■c »"> 


■— 




— 


41 


*■ W5 


• a. 


*# 


■*» 


© 


e 


Vegeta 
Type,l 


0/ 

•c 


E 

o 


■ 

•c 


■ 

•c 


j I 


It 


4, oi 


B 


•£ fa 
C oi 


M 


■s 1 


- 


'3 

E 
0. 


Prai 
Bay 


1 s 


2 2 

a. a. 


1 

a. 


'3 1 


ll 

CM 0. 


£ £ 

JS o 

<z> lb 



o 

CN 



«1 <* 

2 * 



d 

CN 



m 



3 a 



en 



o 



■a 

£ 

JS 

c« 

e 

OI 
1 

o> 

a. 



i 
£ 



519 



< 



* 



•a — 
i 

r> 
K 



& 






1 tf 

Zl8 

< ? ri 

E 2 

BO 



/ 



M 

sis; 

*s "2 A 

£ a 
| - 

CD 






* M 
© e 
Z T C 

. ■ in 

•SE* 



< -5 S 



* 



T3 
<D 
3 
fl 

C 

o 



3 



I e 

< 1 >/> 

■ E 2 



M 

■SS; 

e a 

•a ^ 
is «r 

m a 



oo 



00 

© 






Ti- 
en 



© 



o 
© 



cn 



© 






so 



00 

o 



© 






© 



o 
o 



O 



00 

© 



CN 

o 



m 

o 



CN 

© 



cn 



© 



rs 



cn 
cn 



© 



o 



o 
o 



«-> 






SO 
© 



SO 
csi 



SO 
cn 



OS 



cn 
© 



CN 

o 



2 

5 

e 

5 I 



tn 








1 


•c 


e 


§ 




e 


e 


e 


u a 

J3 A 


e 
•S "C 


1 

a. 

S 

.a 


B. 
O 

e 
^ - 


B 




■ 

"O O 

S U. 


■ 
i 

£ 


nd Pine t 
b-Bay 


g g 


11 


5 


2 1 


B 


o 


J « 


■3 E 


JS JG 


•= - 


j= 


- > 


n 




&4J 


3? 4 


- j= 


C« «3 


en e> 


(/> 


a. 


a. 


3 SS 


3 81 



e ■> 

!! 

3 9! 



if B8 



■ 

"a. 

S 



U a. 



520 



-o 

D 

3 

fi 

'•^ 

C 

o 
o 

I 



u 

a 
H 



H 
















rea 

urin 

990 












00 




of A 
edd 

77-1 


O 




O 


O 




00 


(N 


# |2 
















CO 
















*r m 
















o e 
















of Area IN 
irned duri 
1977-1990 












Tj- 


00 


m 










rn 


00 


o 
© 




O 


o 




OS 


<N 


* J 
















M 
















_ B 
















Area 
duri 
1977 


























i— < 


SO 












SO 


1— 1 


*s "2 « 


o 




o 


o 




u-> 


SO 














00 


-*• 


GO 
















*» wd 
















• e 
















■ 1 £ 












o 


Os 


% of Are 

Burned d 

1952-1! 












SO 


SO 


o 




o 


o 




o 


Os 


M 
















1 -e Ci 












Tfr 


o 


<; -o S 












i— ■ * 


r~ 


•s "° j£ 

■fin 


o 




o 


o 








55 £ £ 












00 


<T> 


s 
















GO 
















*■ M 
















J5 s 
















* t e 
















■ is 












r-- 


SO 


% of Are 

Burned d 

1855-1! 


(N 




ml 






r~ 


O 


o 
o 




o 


o 




00 


00 

m 


4) 
















M 
















IE 


e 


«> 
•c 


2 


e 

41 


- 


■ 


< 


08 — " 


e 
£ 

■o 


1 

e 


1 

■o a 


B 

£ 

■o 




■ 

1 

< 


i 


M C 


B 
1 


a- 2 


B 


B 
4> 


3 


1 


•> »-. 


"EL 


"c. 


e. 


e 


1 2 


> H 


s 


GA 


3 st 


a o 


H 





r^ 




S 




ON 




^1 




#* 




>0 




*r> 




ON 




•— < 




1 

C4 


M 


WN 


< * "T 


On 

•— < 


| 


<- - © 
© 5 on 


1 


*{S 


T3 


■ 



NO 

I 

</1 




522 



3 
C 

1 






en 








































• 


«r« 










































Are 
dui 
-199 


© 




t-~ 




o 




r*1 




** 




vq 




m 




o 






© 




o 


•— s p 


o 




~~ 








O 




© 




~~ 




^™ 










i/"l 






S? E 2 

° 3 










































■ 










































*■ m 










































5 1 










































08 3 S 










































% of Are 

Burned d 

1990-1! 


r*1 
o 
d 




o 




o 




© 




o 




© 














to 




o 


M 










































rea 

urin 

990 










































< -o - 


o 




VD 




o 




fS 




v> 




— ; 




— j 




o 






0-> 




o 


<— ^ r- 






»-H 








© 




d 




•ri 




© 




«-i 






<N 






o t. r- 










































* |2 










































00 










































— M 










































J 1 




















































































% of Are 

Burned d 

1977-1 


© 




(*l 




o 




<s 




© 




~" 




© 




1- 






iri 




o 


M 










































1 










































9 'C f) 










































% of Are 
umed dui 
1952-195 


o 
© 








o 




©' 




•/I 
d 




r- 

© 




Tf 




\6 










o 










































CO 










































*• M 










































J c 










































fll 










































% of Are 

Burned d 

1952-1! 


o 
© 




CO 




o 




3 

© 




d 




6 














ro 




o 


4) 

li 

S2 


e 

m 
e 








e 


u 


e 




1 




e 


£ 


© 


S 


e 
-*- 


l 

E 

3 




i 




£ i 

■ B 
t JL 


e t 


*- 


•- 




■— 


a 


•*— 




•*- 




«- 


«« 


, 0= 


». 




••— 




o 2 

•■e 


s 


3 


c 




I 


08 


w 

£ 




en 




8 


£ 


£ 


08 j> 


£ 


u 

1 




£ 




88 «r 


o 


"9 


<u 




© 


e 


4» 


e 




e 


"O 


o 


EC » 


e 


V5 




o 




*> T3 


Veget 
Typi 


fa. 


3 
3 

s 

o 


Eft 


! 

e 
fa. 


fa. 


e 
a 
O 


fa. 

*•* 
Ov 


•c 

J 

6. 


fa. 
* 


E 

en 


fa. 


3 
C8 

a 
5 


fa. 


1 £ 

= - 

o u 


fa. 


£ 

a. 

5 


s 


fa. 


E 

-. 
en 


j! 



523 



M 

E s i 

< -o -. 

W. TS © 

e jj K 

E £ 

s 

■ 



5? 



o 
Z 

< 
i 
* 



M 

II 

s 

aa 



til 

i- u t-- 
© ■ i- 

E £ 

9 
SB 



5? 



5? 



M 

B 

■o 2 

[S 

to 



M 

B 

< * £ 

•a ' 






*o 


Ov 


o 


VO 



© 



c i 



9 s 3 — 



^ 



T3 

■u 
3 
S 

c 
o 
u 

I 
I 

VO 

u 

•s 



s 
bs 



M 

B 

•r </a 

■§2 
E 2 

SB 



M O 
B C 
« ©\ 

•« 

a, p. 



r-4 



ON 



o 



00 

o 



00 









© ©' 



00 

©' 



00 



o 



00 






o 

00 



o 



o 
o 



o 



© 






o 



C) 


**! 


o 
d 


oo 











00 






e 



(/) 



2 

e 



B 
V 

a 
O 

© 

x '- 

li 



•c 
1 



x 
£ 

JS 



JS 

E 

x 
en 

B 

£ 
1 



I 

"a. 

B 

e 
-— 

2 B 



I 

E > 

- « 

t I 

2 s-s 

JS « X 



c 
II 





■ 






fe 


e 


e 


eg 






— •- 


BB 


hi 


>> 

■ 


£~ 


e 


°? i 


BB 


f? 


JS JS 


X = 


■e ~ 


c E 


5 £ 


S t 


il 


JS JS 


J A 


x o 


<Z) (Z> 


(/> O 


K b. 


v) O 



524 



M 

£ is; 

60 




o 
Z 

■ 



g*s 



x 



4> 
3 
C 

c 
o 
u 



1 



3 ~ 

as 



*8 



I 

U 

e 
o 

5 
9* 



41 

a. 
5 



© 
©' 



o 



© 



00 

© 



o 



o 
o 



o 



©' 



o 



o 



©' 



o 



<s 

© 



H 



00 

© 



o 



5? 

ii 
11 

c« a. 



ea 

i 

2 E 



e 
— « 

ea * 

Si 

II 



s s 

A B 52 

■S i 8 
R U U 



I 

c E 
1 I 
£« 

"P | 

5 a 5 
M f+M 

V) U CW 



1 
£ 

en 

e 






OS 

ea 



u 

S 
9S 



si 

es fa 

ea s 



ll 



E fa 

a. £ 

>> as 

f! 

5! O 



a 
U 

II 



525 





T3 
V 



O 

u 

I 
I 



■s 



1 Eg 

< ; 7 

j? It 

aa 



C5 3 X 

o ■= _ 

1 Eg 



M 

e 

1 "2 d 

S? g 25 

aa 



ill 

< 1 ^ 

■ E 2 



o 



o 



o 
o 



©' 



© 



o 



s 









o 



s 



I- 

if 
ft 



O 



o 
o 



o 



o 
d 






o 



© 



E 

a 
U 

£ 



1/5 C/3 



e 

■ 

■ 

- S 

u £ 

21 
El 



£ 5 
tfrl 

2 i Sr 



e 




e 
















*» 




• 














■ 


i e 

U A 




■ 
| 


E * 




4« 




rub- 
pen 




5 2 




O.JO 


- ffi 






£ O 




-= a. 




u i 


9 e 




1 2 




! 5 




<?2 


ru ta- 
pirs 


E 


rub- 
y-Sh 


■ j 


e 


(A «. 






- 


!j 


P* '3 


■O 


- "3 


£ 




*- 


°>"3 


■= ,*? 


« 


JS " 


r^ £ 


■~ 


r*> " 


e 


<« o 


en 


&n aa 


<J 0. 


- 


U a. 


b 


u £ 


U £ 



526 






u 

a 
c 

X 

c 

o 
o 
i 

i 







































































, 














of A 
edd 


O 




o 






© 


00 


rs 


o 




r-> 


o 


— 


O 




©' 






©' 


fi 


fi 






©' 




©' 


* |£ 




























CO 




























-z M 




























i C 




























Zij 




























■ Is 




























% of Are 

Burned d 

1990-1 


©' 




© 
©' 






© 


<•» 




© 




© 


© 


© 


M 




























- 1 -. 




























sis 




























of A 
edd 

77-1' 


o 




© 
d 






© 
©' 


r«"> 

© 
©' 


© 


© 




© 


© 


© 


* |2 




























so 




























— M 




























1 e 


















- 










% of Are 

Burned d 

1977-1' 


©' 




o 
© 






© 


00 


© 


© 




© 


© 


© 


M 




























I 




























n'Cft 




























Els 




























%ofA 
timed d 
1952-1! 


ci 




o 
©' 






© 






© 




©' 


© 


© 




























■ 




























— M 




























3 1 




























« - On 




























%of Are 

Burned d 

1952-1 


d 




© 






vq 


«*i 


©' 


© 




©' 


© 


© 


it 

OS On 


■ 

X 


E 


1 


c 




i i 
-° E 
E = 9 


-1 


■* =L 


e 




e 


e 


e 
*- 


ri 


S 


5 


a. 




E £,, 


E £ 






Jl 

jj! 


fl 

E I 
9 Jr 

m O 


■ 

•c 

I 


Vegetation 
Type, 197 


■a 

1 

■ 


a) 

e 
-*- 

1 


Z 


2 
e 






i u ■§ 

«? e E 
A «• J 

e,'3 g 




s 

« 

■- 
3 


ll 


O. 


1 


4. 

e 


U C CB 5> 


U a. <« 


2 1 

en eo 


6 * 


x a. 






527 






s 

1 

o 
o 

i 



<n 



•s 



M 

< f T 

<- "0 O 

■ 



« = §; 

< 1? © 

511 

5?M 



M 

■ f 1 

£ 5 £ 
<«- 

<■ K t* 

v? F *> 

ft** s w 



J I 

•s it 



M 

■ 

1 "2 ri 

aa 



ac|s 

* IS 

5? £ 



■ 
MO 

1?. 

p 



© 



*© 



<N — 

d © 



d 



oo 

d 



© 



d 



o 



© 
d 



d 



d 



d 



d 



© 

© 



© — i 



©' 



ON 

d 



© 



© 
© 






© 
© 



oo t-» 

v* ©' 







528 



I 



! 

1 



M 


























rea 

urin 

993 


























4 "O — 


o 






f> 


« 


o 


© 


© 


© 


© 


© 


© 


•s 7 © 
e 1 ov 








^ 


d 
















S? E 5 

• 3 


























■ 


























— M 


























• C 

Zt5 


























8 J R 


























c 2 7 


r- 






ro 


IS 


o 


© 


© 


© 


© 


© 


© 


% of A 

Burnet 

1990 


© 






H 


(•» 
















M 


























rea 

urin 

990 


















































< ■» 7 


o 






o 


s 

d 


o 


© 


© 


© 


© 


o 


© 


j? !§ 


























■ 


























- M 


























5 1 


























« 3 OS 


























% of Are 

Burned d 

1977-1 


r«1 

© 






o 




o 


© 


© 


© 


© 


© 


o 


M 


























e 


























4rea 
duri 
1955 


v"> 






*o 


vO 
















%of, 
urned 
1952- 


O 






H 


H 


© 


© 


© 


© 


© 


© 


© 


























05 


























a Not 
uring 
»55 


























% of Are 

Burned d 

1952-1' 


00 






r^» 


On 
















© 






<s 


<S 


o 


© 


© 


© 


© 


o 


© 


&© 
§8 


i 






i 

a 

fl 

■ X 

: i i 








c 










Vegetation Ch; 
Type, 1977-1 


E 

3 



e 
— 

•c 

'3 
- 


i 

1 

o. 
;►> 

U 

• 

«! 


5 


i 

00 

o 

• - 

1 E 

- JB 


pen Water to 
are Ground- 
rban 


e 

t- 
« — 

*1 

e ■ 

O £ 


o 

1 

1 >. 

.2 t 


e 

w 

■ 

£ a 

11 


a 
■ 


pen Water to 
pland Pine 


pen Water to 

um-Bay- 

ypress-Shrub 


a. 


03 


</) 


o. W 


a- en 


Ofl3 


O 5- 


O (A 


O S 


O 8 U 



529 



-a 

3 
C 

I 

O 



■a 



M 














rea 

urin 

993 










r- 


rr 


<*- 










<* 


r4 


© 




o 




VO 


VO 


>— ~ c 










m 


PO 


O 4. a 














*js 














ea 














r on 














8 c 

2 * « 














rea 

dm 

■199 


Q 




o 






$ 


of A 
irned 
1990 










CI 


s> 


,? £ 














on 














f Area 
d durin 
7-1990 














o 




o 




O 
00 
00 


00 


C <U I— 










*— ' 


^* 


* ES 














ea 














- u 














JS b 














Zfo 














8-3 1 










m 


■<r 












00 


00 


o 




o 




in 


00 


■ E 2 










o 


o 


M 














e 














«T») 














% of Art 
umed du 
1952-195 










t» 


o 


o 




o 






i 










*~ t 


■ 


09 














«• 01 














5 I 














Z T K> 














% of Area 

Burned dui 

1952-195 


© 
© 




© 




o 
CM 
(S 

o 


00 

v> 


& = 














e c- 












« 




e 


1 


e 




CB 


t 


etation CI 
pe, 1977- 


k 


E 
s 

e 

■ 

91 

■ 


k 
08 


5 


JS 


< 

•e 
& 
t 


e 


£ -S 


e 


(W 


75 
e 
H 


& 


ge 


Ok K m 

OUk 


O CO 


11 



530 
had been persistent upland pine or had changed from shrub to bay-shrub or shrub to 
cypress-blackgum-shrub during 1977-1990. These types of changes also occurred in 
areas that did not burn during that period. 
Wildfire and Logging 

During 1895-1942 nearly all of the marketable cypress, pine, bay, and blackgum 
were removed from the swamp interior and near perimeter uplands within the current 
refuge boundary (Table 5-7). Approximately 32682 ha (26%) of the swamp were logged. 

Table 5-7. Composition of logging harvest during 1909-1927, from Hopkins (1947) and 
Izlar(1984). 



Species 


Volume (cubic meters) 


Pond Cypress 


1,842,771 


Slash, Pond, and Longleaf Pine 


555,651 


Swamp Red Bay 


2,552 


Swamp Blackgum 


284 


Red Maple" 


199 


Live Oak b 


199 


White (Sweet) Bay 


142 


Sweetgum 


28 



" Acer rubrum 

b Quercus virginiana 

c Liquidambar styraciflua 



During and following the timber harvest, fires burned periodically in the swamp, 
frequently in the logged areas where logging debris had accumulated. During 1855- 



531 
1952, 23.0% of the area burned by wildfires was along previously logged tramlines; 

64.2% of the logged area burned during that period (Table 4-21). The extensive fires of 
1954-1955 burned in 79.8% of the previously logged area, which comprised 19.8% of the 
burned area in 1954-1955. The swamp did not burn extensively during 1956-1980; 
during this time only 1.0% of the swamp interior burned, and 43.0% of this area was 
previously logged. Of the area that burned in 1990-1993, 8.8% was previously logged; 
5.7% of the logged areas burned during this period. 

Before the area was logged, approximately 25% (1.94xl0 8 kg) of the total 
standing fuel load in Okefenokee Swamp was in the tramline areas, which made up 26% 
of the swamp area (Table 4-19). Wet forest communities comprised 95% of this volume, 
and most (89%) of this was pond cypress. By 1990 the area previously logged contained 
approximately 1.3x10" kg of fuel; 80% of this was wet forest communities, and cypress 
trees made up 57% of the standing fuel. Non-cypress tree species contribution to the 
tramline area fuel load increased from 8% in 1855 to 23% in 1990. Shrub composition 
in the tramlines changed during 1855-1990; tramline area fuels were 2% shrub in 1855, 
8% in 1977, and 13% in 1990. These trends follow the vegetation changes occurring 
outside of the logged areas, suggesting they are not exclusively the result of logging. 
Recurrence of Fires 

Large fires occur periodically in the swamp, but the majority of fires are small 
and burn less than 1% (1600 ha) of the swamp. Of the area burned in wildfires during 
1855-1952 (particularly during the 1931-1932 fires), 74.1% burned again in 1952-1977, 



532 
and most of this occurred during the 1954-1955 fires (77.2%) (Table 5-8). Nearly all of 

the area that burned in 1977-1990 had burned in 1952-1955 (99.7%), and 71.5% had 

burned during 1855-1952. Areas that burned in 1990-1993 also burned in 1952-1955 

(99.4%), and 41.5% burned in 1855-1952; only 1.8% burned during 1977-1990. There 

appears to be a 75-100 year burning return frequency for most of the swamp, and this 75- 

100 year fire consumes roughly 75% of the swamp. Fire recurrence is slight after this 

fire for 30-50 years, but if fire is permitted to occur, the burns will remain small 

(< 10,000 ha) and less severe until appropriate weather conditions occur (drought) and the 

fuel load, which has accumulated for 75-100 years since the last large fire, is sufficient to 

perpetuate a large fire. Small fuel loads or non-drought conditions may carry less intense 

fires throughout the swamp, but changes in direction of vegetation succession are 

dependent on severe, hot fires which are more likely in drought conditions after fuel has 

accumulated. It is probably the combination of 30-year drought cycles and 75-100 year 

fuel accumulations that eventually converge to produce a large, hot fire that alters the 

landscape structure and creates the "moving mosaic" of communities within the swamp. 

Fire Occurr ence and Water Levels 

Throughout 1941-1993 water levels at SCFSP and SCRA were negatively 
correlated with wildfire number and size (Table 5-9). Conditions under which wildfires 
occur in the swamp were fairly consistent during 1941-1993. There was a constant 
number of small fires occurring primarily during the summer months of 1941-1993. This 
corresponded to water level conditions which were dropping or low (Figure 5-13). The 



533 



•5 e Z? 

* i 3 

w S 9 

P <S ON 

U — r- 
< 



m 

•o c © 

** •= o\ 

* It 



4> 



OS 



3" 



S 






•B ©N 
** i 

C? c ffi 

S? .= vi 

W™ ©N 

es 5 rt 

<u e 

'- u 

< 3 

a 



es 

** .5 IT 



>r> 


«* 


ON 


^f 


ON 
ON 


ON 
ON 



00 






On 
On 



00 
00 



NO 



O 

o 



00 



m 

od 






§ 






*r\ 



Cn| 

t-»' 



? 



On 
00 



On" 



> t 

fa. *" 



• Jt «=t ^ J C « .S w « 



« 5 if, 
xi e on 

** ™ *r 
Cng 



© 

"m = C 



« .5 w Ijt 

«• 3 M «3ti ?? s 6 







534 


Table 5-9. Spearman rank order correlation comparisons (t m P) of wildfire size 


, water 


depths, and wildfire cause, for wildfires occurring during 1941-1993. 






SCRA Total Number of 


Number of 


Control 


Interval and 


Water Burn Number of Lightning 


Incendiary 


Burn 


Parameter 


Depth (m) Area (ha) Wildfires Ignitions 


Ignitions 


Area (ha) 


1941-1993 








SCFSP Water 


0.8561 -0.1559 -0.1594 -0.1477 


-0.0568 


0.1858 


Depth (m) 


0.0001 0.0001 0.0001 0.0001 


0.0428 


0.0001 


SCRA Water 


-0.1542 -0.1528 -0.1224 


-0.0750 


0.0894 


Depth (m) 


0.0001 0.0001 0.0001 


0.0074 


0.0014 


Burn Area (ha) 


0.9297 0.6020 


0.6205 


0.0341 




0.0001 0.0001 


0.0001 


0.2238 


Total Number 


0.6352 


0.6382 


0.0167 


of Wildfires 


0.0001 


0.0001 


0.5516 


Number of 




0.0187 


-0.0448 


Lightning 
Ignitions 




0.5063 


0.1104 


Number of 






0.0830 


Incendiary 
Ignitions 






0.0030 


1941-1959 








SCFSP Water 


0.8929 -0.1874 -0.1801 -0.1459 


-0.1090 


-0.0399 


Depth (m) 


0.0001 0.0001 0.0001 0.0018 


0.0199 


0.3949 


SCRA Water 


-0.1975 -0.1979 -0.1495 


-0.1091 


-0.0319 


Depth ( m ) 


0.0001 0.0001 0.0014 


0.0198 


0.4966 


Burn Area (ha) 


0.8373 0.3912 


0.6213 


-0.0220 




0.0001 0.0001 


0.0001 


0.6400 


Total Number 
of Wildfires 


0.5191 
0.0001 


0.6250 
0.0001 


-0.0262 
0.5764 


Number of 

Lightning 

Ignitions 




-0.0333 
0.4779 


-0.0134 
0.7757 


Number of 
Incendiary 
Ignitions 






-0.0165 
0.7254 



Table 5-9-continued. 



535 





SCRA 




Total 


Number of 


Number of 


Control 


Interval and 


Water 


Burn 


Number of 


Lightning 


Incendiary 


Burn 


Parameter 


Depth (m) 


Area (ha) 


Wildfires 


Ignitions 


Ignitions 


Area (ha) 


1960-1993 














SCFSP Water 


0.8697 


-0.1754 


-0.1663 


-0.1827 


-0.0428 


0.1409 


Depth (m) 


0.0001 


0.0001 


0.0001 


0.0001 


0.2222 


0.0001 


SCRA Water 




-0.1492 


-0.1364 


-0.1271 


-0.0636 


0.0800 


Depth (m) 




0.0001 


0.0001 


0.0003 


0.0696 


0.0222 


Burn Area (ha) 






0.9771 


0.6801 


0.6226 


0.0346 








0.0001 


0.0001 


0.0001 


0.3241 


Total Number 
of Wildfires 








0.6908 
0.0001 


0.6453 
0.0001 


0.0244 
0.4868 


Number of 

Lightning 

Ignitions 










0.0394 
0.2612 


-0.0678 
0.0531 


Number of 
Incendiary 
Ignitions 












0.1075 
0.0021 



536 



seJtfPHM # |B*oi A|iftuow 



(0 





1 


! 1 

— ^r • 

^ r >..- • 






C___^}- • 












.sr^r 




• 


<* < • 

~5c i 


• 




x* ••■"■''" 
: — <f ::::: * 

^* 



i 



2 



in 

% 



a 



i 



CM 

i 



s 



| 



< 

O 
(0 



© 

u 

I 

I 

♦ 

£ 
a 
a> 
o 

L. 

3 
a 

$ 

a 

CO 
U. 

o 

CO 



(ui) ifldaa je;e/v\ 



537 



saJ!iPl!M # moi Ajiftuow 



<o 



W 



eo 




0) 

in 

0) 



s 



in 



8 



w Is 

ID TO 

05 « 

T- > 



o 

I 

< 

cr 
O 
eo 



3 

O) 



2 

O) 



CM 
W 
0) 



in 

0) 



8 



0) 
I 



| 

o 

1 

5 

Q. 
(0 

u. 
O 
(0 



(m) iftdea je»BM 



§ 



rn 



*r, 



U-i 



538 



S8Jypi!AA # |B»oi Ajinuom 




(ui) iftdea je»BM 



§ 



m 



«n 









539 



S8JUPHM # |e*oi ^liftuow 



co 



CM 




O 


< 


K 


rr 


0> 


o 




(0 




£ 


eo 


a 


r- 


9 




a 




u 




5 


K 
K 


a 


0) 




*- 




co 




r» 




0) 




»• 




in fa 


CO 


k re 


P> 


0) CO 


i. 


r > 












s 






♦ 


co 




K 




CD 




T- 






£ 


CM 


a 


0> 


a 


T* 


u 




0) 




•j 




re 




5 


0) 


1 


f" 


1 



CD 



« 



U) 



IA 



(ui) iftdea je^BM 



? 

■ 

| 

E 



540 



sa JijPMM # |B»01 A|lflUOW 
to CM 



0) 



eo 



• 


I 1 i 


co 




^jP?* 






"""^bT ? '> # 


3 


< 




*^» «J5» 


^ 


tr 




^J5r* ^ 


CM 

i 


o 

(0 








a 






O) 

8 


o 

o 


?*">" 




• #^=^ 


^ 


o 




•5- — r^^~ 


o 


a 




" =c: -^-—S Si ..., , 


0> 
0) 


5 






■ 
S 






;^b=> >• t 


eo 








oo 

0) 






.^ -~^t 


T" 


0) 




*s^_^~~~c~ 


eo « 


C 




__JJFJ^I 


05 © 
ID 


s 

CD 

re 




■*cr '""?z % 


eo 


© 




^*""^t *■** 


o> 


1- 






«j« 


< 




^^=-.;t| 


eo 


♦ 




_J5> £""*'"" • 


^ 


0. 




.^^^ *<- • 


I 


(0 

LL 




jP"^ r-^- 5 "* 


o> 


o 




*— -^ "■"■■"-, 


eo 
eo 


£ 




*^ ""c* 


CO 


a. 






CM 

eo 


u 
CJ 




^**^ 


o> 


t 




^~ • 


oo 


5 

1 






8 






J^* sr- 5 * • 


IF 


1 




:sr .c" • 


o 




! 


1 1 


eo 

8 





CM 



IO 



IO 



(ui) mdea je*eM 



1 

c 

O 

o 



m 



*o 



u. 



541 
largest fires occurred when water levels were low or were increasing following a low 
period (Figure 5-14). During 1 94 1 - 1 959 the number of fires with incendiary or lightning 
ignitions was also correlated negatively with water level; number of incendiary fires was 
not significantly related to water level conditions during 1960-1993 (Table 5-9). 

Most prescribed burns were carried out in upland areas during December-March, 
when water levels are usually high (Figure 5-15). Although a few of the incendiary fires 
originated as prescribed burns and accidentally spread, most of the prescribed burns were 
well-controlled. There has been an increase in prescribed burning area during the past 
two decades; acreage burned in prescribed fires exceeded that consumed in wildfires 
during 1970-1979 and 1980-1989 (Figure 5-16). However, the prescribed burning 
program has not replaced or displaced swamp wildfires. Most of the controlled burning 
was on interior islands and perimeter fire compartments composed primarily of upland 
and wet pine and upland shrub communities, rather than in the interior wetland types. 
Most of the wildfires occur during the summer months, when water levels are low; the 
refuge does little prescribed burning during this season due in part to fire hazard 
conditions which increase in the low water period. However, changes are occurring in 
the prescribed burning program to include more summer burns. 
Fire and the Suwannee River Sill 

During 1855-1993 there were 26 wildfires that originated in or burned into the 
area of the Suwannee River floodplain currently affected by the sill (Figure 5-17). 
Thirteen of these fires occurred after the sill was constructed (Figure 3-24). Ignition 



542 



S8JUPHM Aq peiung (eq) eajy mo± Aimuo^ 




f 

o 
O 



5 
< 
or 
o 

CO 



3 
CD 

R 

£ 

< 

1 



f 



5 

Ol 
CO 

u. 
o 

CO 



- 



J 



(ui) iftdea je^BAA 



■a"! 



sajijpuM Aq peiung (en) eejy \e\o± A\muo\fi 
spuesnom 



o 

CM 



o 
o 



o 



© 

CO 



o 



o 

CN 



I 1 


III! 






S 




— — -**— -2* 


s 

0) 


a 
o 

Q 








•"* _./'■ 


^ 


k 




s — ** 




5 




.—'■'' 




(8 




~ e= ^? 1 


00 

in 


$ 




^^ "*:i 


o> 


< 




^^„ <» 




a 






r» 


CO 




""^7 J 


k w 

' o> 






^^p "'"";;;::::= 


^ 














*"*- -■'*"' 


CD 


"O 




~==i^L.--''~ 


8 


a> 
E 








3 




J. 2 


\ 


CD 


• 


■^# 


in u 

m to 


a 




• i 


o> <o 

1 ** *■ 


2 

< 






_ 




v^*""^* " 


's 


5 
o 




^~~> — ::• 


B 


t- 






T" 


♦ 


'" — --^ ...-J." 1 




— -> S> f s 


S 






© 




^_s^ --■*• i 




o 




^^-r^--"' i 


' CM 


L. 




^^ """*.> 


in 


5 




ZH^* ,.;: : < 


o> 

1 


re 
5 








0. 




___^5-^ v ' 


in 

0> 


(0 

u. 




'•'"'"■ay 


T" 


o 




<^§££ 




CO 




"~> ::: ':.> < 


0> 


1 




i 


! 1 


^ 





in 



(ui) iftdea jeieAA 



544 



sajypuM Aq paujng (eq) ea.iv l<^oi A|q}uoyy 






< 


8 


O 


8 


CO 


T- 






I 


00 


o 


IB 


Q 


0) 


u 




I 


r«. 


$ 


CD 




o> 




^ 




09 
0) 


o 
E 




3 




m 


«D n 
CO 0) 


cs 
9 


** > 


< 








3 


8 

8 


* 


^ 


♦ 


CO 
09 


£ 


0) 


Q. 


<P 


03 




Q 




L. 


CM 


0> 


09 


♦* 


0) 


5 


T- 


£ 




O. 




CO 


09 


U. 


9 


O 


*" 


CO 


o 


1 


09 




0) 





(ui) iftdea j8»bm 



f 



V) 



s9Jijpi!M Aq peujng (eg) eajy mo± Aimuoytj 



o p o 

o o 

01 00 f- 



o 
o 
to 



© © o 

o o o 

W ^ CO 



o 
o 

CM 




0> 
0) 



co 

0> 



0> 



CD 
0) 



• in b 

en a> 



s 



CO 






0) 



CM 



W 



(ai) iftdea j©»bm 



















545 














< 




K 




O 




0) 




f 




s 




k. 




Q) 




IS 




§ 




1 




] 




E 




3 




CD 




(0 




e 




w 




< 




1 




o 




H 




♦ 








f 




O 




Q 




t- 




a 




a 




5 




1 






•d 




i> 






3 






H 






"^ 






c 






o 






o 






■4 






1 






n 






E 






3 


1 




• *•* 






U. 







546 



saJiiPHM *q paujng (ei|) eejy |B*oi Aiqjuo w 
spuesnom 

o 

T 00 03 *■ N O 




(Ul) IftdBQ J8JBAA 



3 

c 
o 



I 



547 




(spuesnoqj.) 
S8J| j paquosajd jo (eg) eajy |ejo± 



m 
a* 

I 

en 

o 

c 






T3 

C 

I 

u 






1 

o 

i 

I 

o 

c 

O 
.*! 

o 

o 

03 



I 



E 

3 



548 



CO 


0) 


L. 


< 


c 


C 


« 5 


0) CD 


<£ 


C U. 


c 


3 "D 


OQ Q) 


© - 


2 W 


Idfi 
esc 


s& 




1 



5 

^ 

1 



*fl 



<?/ 



'%, 



*z 



<fc 



»A 



% 



»/ 



*« 



/^ 






e * 



«/ 



CO 

> 
L 



^ 



*A 



°S t 



*l 



*k 



*/ 



Vfl 



<>l 



%i 



*/ 



*3 



*Z 



ooooooooo 

tD^tCMOCOW^-CM 



(spuesnom) 
sajj j Aq peujng (eij) eajv |ejoi 



1 



I 

1 

— 

o 



,2 «*> 

o <* 

H £ 

■ 00 

<U OX) 

H C 

a s 



549 













«4- 

o 

■♦w 

w 



On 
O 

i 

O 
vO 



On 

o> 



00 
DO 

c 



I 



ca 

c 
.o 
-^ 

z 

n> 
<u 
_*: 
O 

c 

O 

| 



CD 

c 
2 
'■♦2 

S3 
O 

o 



m 
u 



U 

c 

I 

o 
u 

u 

c 



So g> 



550 
source of 20 of these fires was lightning; most (17) of these fires began in June- 
September. Drought conditions occurred when 17 of the wildfires burned in the sill area; 
water levels were low (3), average (2), or unknown (4) for the remainder. Eleven fires 
burned in the Cypress Creek watershed during 1 855-1993; 6 occurred after the sill was 
constructed. In contrast to fires in the Suwannee River floodplain affected area, most of 
the Cypress Creek area fires were not lightning-caused. Those that were lightning 
ignitions occurred under low or drought conditions in June-September. 

Wildfires burning into the Suwannee River floodplain area affected by the sill 
were smaller following sill construction (x=122 ha, n=14) than before (x=20246 ha, 
n=13). Although this suggests a decline in fire size due to the sill, that conclusion may 
be unfounded. The 4 largest fires burning into the area before 1959 were ignited outside 
the area of the future sill's influence during drought or low water conditions (Table 5- 
10). Drought or low water conditions occurred when 9 fires burned during 1855-1959 in 
the future sill's impact area. It is unlikely that these fires would have been arrested by 
the sill had it been in place; fires also burned in the area following sill construction when 
water depths dropped to low or drought levels throughout the swamp, and similar 
conditions occurred during the large fires before sill construction. Only 2 wildfires 
occurred in the area during 1941-1993 under average water level conditions; ignition 
sources for both of these fires were incendiary. These fires began to burn in upland areas 
(the Pocket and sill berm), and were extinguished by refuge personnel. 

Wildfires in the Cypress Creek watershed also covered smaller areas following 
sill construction (x=1374 ha, n=6) than before (x=3133 ha, n=4 excluding the large fires 









551 



Table 5-10. Wildfires occurring in the Suwannee River floodplain and Cypress Creek 
watershed areas affected by the sill, water level conditions when the fires ignited, and the 
water level condition that would have been required to create impounded surface water 
and arrest the spread of these fires. 



Fire Date 


General Fire 

Location or 

Ignition 

Point" 


Fire 
Size 
(ha) 


Cause of 
Fire 


General 
Water Level 
Conditions in 

Swamp 


Sill-Affected 
Zone b 


Floodplain 
Area 












1874 


Suwannee 

Canal logging 

Tramline 


480 


lightning 


unknown 


high 


1915 


Sill area 
logging 
tramline 


770 


lightning 


unknown 


drought 


July 1931 


Southwest of 
sill area 


18643 


lightning 


drought 


low 


April 1932 


North half of 

sill area to 

Billy's and 

Floyd's 

Islands 


77364 


incendiary 


drought 


drought 


10 June 1941 


Tip of Pocket 


10 


lightning 


drought 


low 


6 April 1943 


Pocket Area 


36 


lightning 


drought 


drought 


8 June 1945 


Southeast of 

Rowell's 

Island 


30 


lightning 


low 


low 


14 April 1952 


Billy's Island 


61 


incendiary 


very low 


low 


6 July 1954 


Billy's Island 


1405 


lightning 


drought 


low 


19 August 
1954 


Pocket tip 


5 


lightning 


drought 


low 


20 August 
1954 


South end of 
Pocket 


<1 


lightning 


drought 


drought 


4 September 
1954 


West of sill 


8 


lightning 


drought 


low 



Table 5-10-continued. 



552 





General Fire 






General 






Location or 


Fire 




Water Level 


Sill-Affected 


Fire Date 


Ignition 


Size 


Cause of 


Conditions in 


Zone b 




Point* 


(ha) 


Fire 


Swamp 




5 March 1955 


Everything 

south of 

Sapling 

Prairie 


164380 


incendiary 


drought 


drought 


10 May 1963 


Pocket 


4 


unknown 


very low 


drought 


21 March 1968 


Sill dike, Pine 
Island 


89 


incendiary 


average 


drought 


3 October 1983 


Pocket tip 


<1 


incendiary 


average 


low 


6 June 1985 


Pocket 


1174 


lightning 


drought 


drought 


27 June 1988 


Between 

Hickory and 

Palmetto 

Islands 


2 


lightning 


drought 


drought 


10 August 
1989 


NWof 
Floyd's Island 


<1 


lightning 


drought 


high 


10 August 
1989 


NWof 

Floyd's Island 


<1 


lightning 


drought 


high 


1 September 
1990 


West of 
Pocket 


26 


lightning 


drought 


drought 


3 September 
1990 


Pocket tip 


<1 


lightning 


drought 


low 


5 September 
1990 


West of sill 


1 


lightning 


drought 


low 


10 September 
1990 


West of sill 


<1 


lightning 


drought 


low 


8 August 1991 


SE of Floyd's 
Island 


<1 


lightning 


drought 


high 


4 August 1993 


E of Floyd's 
island 


<1 


lightning 


drought 


high 


5 September 
1993 


East of 

Minnie's 

Island 


405 


lightning 


drought 


high 



Table 5-10--continued. 



553 





General Fire 






General 






Location or 


Fire 




Water Level 


Sill-Affected 


Fire Date 


Ignition 


Size 


Cause of 


Conditions in 


Zone b 




Point- 


(ha) 


Fire 


Swamp 




Cypress 
Creek Area 












June 1927 


Cypress 
Creek 


8152 


lightning 


drought 




July 1931 


Cypress 

Creek and 

West 


3743 


lightning 


low 




17 February 
1941 


SW Sapp 
Prairie 


607 


incendiary 


low 




31 October 
1954 


SW Strange 
Island 


30 


incendiary 


drought 




5 March 1955 


throughout 
watershed 


164380 


incendiary 


drought 




27 September 
1980 


SW Sapp 
Prairie 


10 


incendiary 


low 




27 July 1987 


East of 

Cypress 

Creek 


12 


lightning 


average-low 




31 March 1989 


SW Sapp 
Prairie 


2 


incendiary 


average 




10 August 
1989 


Sapp Prairie 


<1 


unknown 


low 




October 1990 


NWSapp 
Prairie 


8217 


lightning 


drought 




27 December 
1990 


SW Sapp 
Prairie 


4 


incendiary 


low-drought 




a See Figure 2-1 1 


for map of place 


: names. 









b Sill Affected Zone refers to the region of the swamp near the sill that is impounded by 
the sill in high, low, and drought water level conditions (Figure 5-18). The Cypress 
Creek watershed is included in entirety for all water levels. 



554 
of March 1955). Only one fire, of incendiary origin, occurred during average water level 

conditions; drought or low water conditions occurred when the other fires ignited. The 
incendiary fires which burned in the area with and without the sill also occurred during 
low or drought conditions. This area retains water during periods of abundant 
precipitation, but when precipitation is limited, conditions approach pre-sill levels. 
During high water level conditions since sill construction, this area may actually de- 
water more rapidly than prior to sill construction. The greater difference between water 
levels in the creek watershed and the river below the impounded water at the sill 
facilitates more rapid draining of the creek (see Chapter 3). This difference and the rate 
of creek drainage decreases with declining swamp water levels. Most wildfires are 
ignited by lightning strikes during June-September, when water levels are falling due to 
high levels of evapotranspiration. Thus the sill impoundment effects in this area occur 
primarily when wildfire potential is low. 

Although the sill may provide fire protection under a limited range of water level 
conditions, its performance is not perfect. Areas in the Suwannee River floodplain and 
Cypress Creek watershed impounded by the sill during various water level conditions are 
delineated in Figure 3-24. Following sill construction four fires occurred during drought 
in the floodplain region that probably was impounding water, but only within the river 
and stream beds. Three wildfires occurred during drought in the floodplain area that had 
the potential to contain some impounded water during low water conditions, but not 
necessarily during drought conditions. In the area impounded in the floodplain only 
under average to high water level conditions, one fire occurred during drought. One fire 



555 
occurred under average water level conditions in the area that is impounded during low 
water conditions. These fires were extinguished before burning more than 400 ha. 
Because of generally low water level conditions when these fires occurred, it is possible 
that they would have had a much greater spatial extent if permitted to burn, in spite of 
the presence of the sill, because the extent of the impounded water was most likely 
limited to the river and stream beds. Wildfires in the Cypress Creek watershed also 
burned during low water level conditions; although fire suppression efforts controlled 
several of these fires, the largest burned until extinguished by precipitation. 

Discussion 

Determining the impacts of the Suwannee River sill on the Okefenokee Swamp 
hydrology and vegetation requires that the "natural" successional sequences and 
disturbance patterns be recognized, as well as the effects of man-induced processes that 
have affected the landscape in the past, such as logging, draining, peat mining, and fire 
control. Only when all of these effects are assessed simultaneously can the impacts of 
the sill on the current hydrologic environment and vegetation communities be assessed. 
There have been changes in the Okefenokee Swamp vegetation composition during the 
past 40 years that indicate the system is becoming more forested than in its recent past; 
examination of the history of these trends suggests that logging and fire suppression have 
contributed to these changes. 



556 
There were 2 periods since the early 1900s when most of the swamp was burned 
by wildfires. Both of these fire periods included lightning and incendiary fires, and 
occurred during droughts. However, even though most of the swamp was burned, the 
vegetation composition changed only in a small proportion of the burned area (Hamilton 
1984, 1982, this study Chapter 4). This finding suggests that the fires were not severe, 
although they were extensive. Evidence of extensive, historic fires exists in the peat; 
during the past several thousand years, forest communities existed where prairies occur 
today and prairies existed in currently forested areas (Cohen et al. 1984, Cohen 1975, 
1974, 1973a, 1973b, Fearn and Cohen 1984, Rich 1984a, 1984b, 1979). Response to the 
1954-1955 fires indicate that a forested area requires intensive burning repeatedly over a 
few to several years to result in a prairie, or a fire needs to be hot enough to remove peat 
and kill roots buried in the peat or underlying sand to revert a shrub or forest stand after a 
single fire to prairie or open water (Hamilton 1984, 1982, Cypert 1973, 1961). The fact 
that this kind of change occurred in less than 5% of the swamp after the 1954-1955 fires 
indicates that the area has not been recently affected by a wildfire that was severe enough 
to alter succession for more than a few years. Alternatively, the extent of severe fires 
may be only local, but if they occur frequently, they may ultimately create more structure 
and texture in the landscape than infrequent, extensive, but low intensity fires. 

The change in the annual peak wildfire frequency from March-July to June- 
August, and change in peak burn area from July-November to June-August, may be 
attributed to several factors. The shift in fire season may be an artifact of better reporting 
of lightning strike fires under the recent (since 1974) fire management program. There is 



557 
also the possibility that the accumulation of fuel has reached levels that will easily carry 
wildfire if an ignition occurs when water levels are low, which occurs primarily in the 
late spring and late summer months. Late summer also corresponds to the peak in 
lightning strikes. These fires have the potential to get large and severe if water levels are 
low (Yin 1993), since the approaching fall is usually accompanied by decreasing 
precipitation and therefore lower water levels. It is these fires that would probably be the 
most effective in maintaining or changing the swamp landscape, since they accompany 
seasonal drought. 

The prescribed burning program is concentrated in the winter months, when 
ignitions of wildfires are low, and is focused on the perimeter and interior upland 
communities. These areas support slash and longleaf pine communities that benefit from 
the frequent burns, but they represent a small portion (8%) of the total refuge area. In the 
swamp interior the threat of damage to refuge structures, perimeter private property, 
increased fire suppression costs as the fire grows, and danger to visitors restricts the use 
of prescribed fire and prompts suppression of wildfires while they are still small and 
controllable, frequently when allowing them to burn would be most beneficial to the 
swamp landscape. The expense of managing and fighting a wildfire grows exponentially 
with its size; this financial burden must be considered in wildfire management, so that 
the decisions made to extinguish interior wildfires and restrict prescribed burning to 
upland areas and wet seasons are not necessarily advantageous to the swamp ecosystem. 

The frequency of wildfires has apparently increased during the past 15 years to 
levels higher than during the past century, yet they are much smaller in size. Fire 



558 
detection and suppression techniques have improved during this period, so that more 
small fires are detected, and fires can be controlled while they are small. However, there 
has also been an increase in forest fuel loads in the swamp during the past 20 years, 
approaching pre-logging levels. The efficient fire control is presently restricting fire size, 
but may also permit accumulation of fuels that will carry an extensive, severe, 
uncontrollable fire with an ignition during an extreme drought period. The apparent 
cycle of approximately 30-50 years for extensive fires accompanying droughts may 
actually only be for fires that reduce the surface fuels slightly but do not burn into the 
peat, or decrease the possibility of another more severe fire from occurring with 
appropriate conditions. This is evident from the multiple "sweeps" of the 1931-1932 and 
1954-1955 fires that burned repeatedly over the same areas. Although both of these fires 
occurred when water levels were low, the resulting vegetation changes were temporary. 
Vegetation was not killed in most areas, although localized mortality did occur 
(Hamilton 1984, 1982, Cypert 1973, 1961). Peat fires burned but were not extensive, and 
the accumulated litter provided fuels for many extensive fires (Hamilton 1984, 1982, 
Cypert 1973, 1961). The types of community-altering burns of past centuries evident in 
peat cores collected by Cohen et al. (1984) were scarce in these fires. This may mean 
that there are superimposed cycles of extensive fires accompanying drought every 30-50 
years that remove a portion of the fuel but do not alter the landscape structure 
appreciably, and severe, extensive fires that occur every few hundred years with more 
severe drought periods, when fuel levels have accumulated in much greater amounts and 
the peat also burns. This cycling agrees with the periodicity proposed by Yin (1993) 



559 
based on size and frequency analysis (without regard to fire locations) of fires in 

Okefenokee Swamp during 1938-1989, and with periodicity of fires proposed by Rykiel 
(1984) based on nutrient and mineral cycling in the swamp. Formation of prairies or 
lakes in shrub and forested areas might then result from these severe, infrequent fires 
(Figure 5-18). 

Logging that occurred in the swamp during 1890-1942 left various scars on the 
landscape. Since logging occurred there have been several fires in and around where 
logging occurred, probably consuming logging debris. Prior to logging most of these 
areas were cypress or pine dominated; 25% of the standing fuel in the swamp before 
logging was in tramline areas. By 1952, the logged areas had changed to dominance by 
shrub species, and accounted for 4.5% of the fuel load. By 1990 wet forest species 
dominated the tramlines and comprised 13.5% of the standing fuel load. As noted in 
Chapter 4, some of the logged areas have returned to a composition probably similar to 
that before logging. This was possible where coppice growth occurred, or where water- 
dispersed seeds were available. However, most of the area from which cypress was 
removed is dominated today by bay and blackgum communities, and cypress is less 
prominent, although in many cases still present. This is not an artifact of data scale; 
similar proportions in the landscape composition exist regardless of the data resolution. 
This alteration in species composition has the potential to affect many aspects of swamp 
ecology, such as hydrologic regimes by modifying evapotranspiration and flow rates, 
wildlife use of these areas, and fire occurrence and behavior. Cypress, bays, and 
blackgum tolerate wildfire to various degrees. They also carry fire differently (Ewel and 



560 



o 

c 
o 

c 

CO 

E 

CD 
Q. 

*U 

C 
CO 



c 
o 

5 

5 

1 
> 

~-> 

u 

Q> 
X ^ 



£ o 

13 



c 

uj r 

£ > 

3 CO 
O 



CD 

LU «- 

£? 

few 



0) 



>» 
u 

c 

CD 

3 



(0 

i § 

«¥"° 

go fr 

— O 3 

« u 9 

§.8 E 

O o « 



a 5 

2 E o 

£ o © 

< S C a 



•s- 



0) 



h ■ 
to c 
co a> 

*■$< 

T- 0) 
* If 

CO o 

>- s 

> u 
CO 3 
CO TJ 



£ (0 



CD 



x a 

a> a> 

I?. 

0) CD 

co t: 
2 I 

C 

E 
3 

2 

TJ « 
C CO 

to a 

— CD d) 

03 .£ "St 

OJJJ E 
® x o 
cr © o 



I 

3 
CO 

| 

o 

"SI 

II 

a A 
to w 



II 

8 m 

CO u 

CO 3 

«- TJ 

5 « 

co "HT 

9 "5 

o to 

■ o 

-O CO 

0) TJ 

Q) V 

to 5 

y f 

CO 0) 



5 

E 

a 

CO 



c 
o 



co s_ J5 

C CO ~ 

o p .9- 

»to 

S O CO 

CO £ C 

co f a 



!? 



c 
o 

co 5 

9 3 

£| 

co" CO 

§§ 

_ to 

{J 

•S X 

5 co 



c 
o 
'55 



co 



CO 



co .= 

3 O 

CO CO 



c 
o 

CO 

i_ 
3 
TJ 
l. 
CO 





5 
o 

-Q 

*j 

CO 
0) 



co >, CO 

l- £ CO 

»oO 

> *■• Q. 
CO CO x 

co a co 



1 8 

CO m 

il 

.3 co 
CO > 

♦J ° 

CO CO 
CO £ 

CO t- 
TJ ^ 



3 

3 "° 

ZC •* J 

■Q CO 

if CO © 

CO > CL 

ra .2 .8 

or co a 



05 £ 
c 2 

O EL 
CO — 
CO CO 

1 o 

* E 
CO •. 

TJ « 

2 TJ 

C 
C 
3 
£1 
■*-• 
CO 
© 

a 
f J 

51 



E 

m 

© 

tj c 

O •; 

a 5 

X *" 
CO CO 

2 9 

12 CO 





TJ 




CI) 


r 


*J 


-»- 


CO 


c 


i. 


o 


3 


E 


4-1 

CO 


■ 


CO 



CD 
O 

CO .-£ 

m C t 
2 O 3 

II 

S TJ 



CO 
<D 

a 



E 

3 
JQ 
*J 
CO 
CO 

a 
-j E 



3 

c 

CD 

o 



i = 

CD C 
CO » 

o 



CD 

"D 

§ 

c? 



CO 

c 

CO 



a 




E 




A 




£ 




00 




o 




ll> 




^ 




o 




c 








u 




-^ 




o 




c 




•^ 




C/1 




d> 




u. 




c 




T3 












? 




TJ 




C 




CO 




^^ 




x: 




HI) 




3 




O 




v- 




■o 




— 




o 




p 








V) 




c 




<i> 




•*^ 




c 




• ^ 




•n 




c 


<U 
61) 




C 


c 


e3 


o; 


J= 


-«-> 


o 


d) 


C 




O 


c 


•*-» 


o 




*-» 


u 


C3 


on 


3 


> 




■*-j 


o 

c 
G 

3 


C 

u 

3 


CT 


,n 


P 


3 


'— 


VI 


E 


Cm 

o 


3 
1> 


>> 


t- 


c 


•a 

'K 

CD 


c 


i 


a 


T3 

C 


>. 


X 


*-» 




c 


oc 


o 


»— * 


X 


t 


<l> 


irv 






•",) 




-c 


cVn 


T3 




C 


Uh 


BJ 



561 
Mistch 1978), an indication of their dependence on fire in maintaining their presence in 
the landscape. The increase of these species and the potential for them to alter 
movement and behavior of fire in the landscape may affect the response to future 
wildfires, and therefore influence the resultant landscape composition and structure. 
These changes appear to be occurring independent of the sill's effects on the 
system. The hydrology model and water level recorder data analyses indicate that the 
sill's primary effects are during high water periods, by extending the hydroperiod and 
increasing inundation depths over roughly 15% of the refuge (see Chapter 3). The 
increase in hydroperiod and water depth depends on the location within this impact area. 
Areas further away are not flooded with as much water, and when water depths are high, 
theses distant areas are flooded at elevated depths for shorter periods than in areas closer 
to the sill. However the extended hydroperiod effects on species distributions are 
probably only occurring within the region bordered by the sill, south Floyd's Prairie, and 
midway to Craven's Hammock. This area has experienced flooding durations 1-4 times 
longer than they would be without the sill; water depths 0.30-1.00 m above pre-sill levels 
(see Chapter 3); and, species composition has changed since the sill was built (see 
Chapter 4). The Cypress Creek watershed has also been affected by the sill; drainage in 
this area may be accelerated during high water levels as the sill impounds water and 
reverses the hydraulic head at the river-creek junction. During average water conditions, 
water levels are higher in the sill's presence than in its absence. The mechanism for this 
change is unclear. Water levels in the Sweetwater Creek drainage, which is closer to the 
sill, should experience a much smaller change with sill removal than the Cypress Creek 



562 
area (Table 3-6). However, at extreme high water levels Sweetwater Creek also 

increases drainage as the hydraulic head reverses (Figure 3-19), as experienced in 
Cypress Creek. Manipulations in the perimeter landscape may also be responsible for 
increasing water levels in the Cypress Creek and Sweetwater basins, by direct 
contribution to the swamp where ditching and clear-cutting have occurred, or by 
increasing water levels between the swamp and the Suwannee River, and therefore 
slowing drainage of the swamp. Beyond these areas there are no substantial effects of 
the sill on water depths or hydroperiods that could be detected with the hydrology model. 
During low water periods, the sill impounds water only in the river and creek beds, and 
to some degree in the floodplain between Billy's Lake and the sill. It is during the low 
water and drought periods that wildfires have been most frequent in this area before and 
with the sill in place. Most of these fires occurred in the swamp interior and were 
lightning-caused. Two recorded fires during the last 30 years near the sill-affected area 
were incendiary and occurred during average water level periods in upland areas 
bordering the swamp; both were extinguished before they entered the swamp interior. 
Fires in the Cypress Creek watershed have occurred throughout the year; most of these 
were not ignited by lightning but were accidental, arson, or escaped prescribed fires. 
However, all of these fires occurred during low or drought conditions and were 
extinguished by fire suppression efforts or precipitation, not by water impounded by the 
sill. Thus the sill does not seem to be arresting the spread or occurrence of wildfires, as 
demonstrated by their continued occurrence since its construction in areas that burned by 
wildfires prior to the sill's construction. 



563 
The management plan of the Okefenokee Swamp ecosystem must consider the 
influences and expressions of past made-made perturbations on the current swamp 
landscape. Because these effects have modified species composition and community 
responses to disturbances, they must be considered when examining the current swamp 
landscape. At a minimum, the variability in vegetation distributions caused by fire and 
drought disturbances must be permitted to occur. The Okefenokee Swamp landscape 
evolved with this variability and will only be maintained with its continued influence. 



CHAPTER 6 

RELATIONSHIPS OF OKEFENOKEE SWAMP VEGETATION DISTRIBUTIONS 

AND THE HYDROLOGIC ENVIRONMENT 



Introduction 

The recognized position of wetlands between terrestrial and aquatic environments 
reflects a gradient of hydrologic conditions that requires the inhabitants endure a variety 
of physiological stresses (Mendelssohn and Burdick 1988). Occurrence and composition 
of specific types of wetlands are predominantly determined by the hydrologic 
environment (Mitsch and Gosselink 1986). The wetland's hydrologic regime (including 
flooding duration, depth, and periodicity) influences species composition by affecting 
nutrient transport and availability, substrate elevation, and substrate organic and 
inorganic composition (Flebbe 1973, Gosselink and Turner 1978, Rykiel 1977). Species' 
tolerances of these conditions, and competitive interactions for available resources while 
enduring these conditions, result in the standing vegetation composition, structure, and 
distribution (van der Valk and Welling 1988). Franz and Bazzaz (1977) suggested that 
life history processes such as timing and means of seed dispersal, germination 
requirements, and seedling growth rates, may be as important if not more important than 
physiological and structural mechanisms of flood tolerance in establishing vegetation and 
succeeding in competitive interactions along a flood gradient. Although average water 



564 









565 
depth might affect species distributions to some degree (Gill 1970, Monk 1966), 

inundation duration and periodicity are the hydrologic signatures of most wetland types 
(Mitsch and Gosselink 1986, Penfound 1952). Deuver (1988) hypothesized that duration 
of inundation was more important than inundation depth in delineating species groups in 
Corkscrew Swamp, Florida. He also found that major community types clustered by 
maximum wet season water depths and hydroperiods. David (1996), Richardson et al. 
(1995), Gunderson (1994), Wood and Tanner (1990), and Loveless (1959) related 
species' occurrences to inundation depth, duration, and frequency in the Florida 
Everglades system. Harms et al. (1980) recorded differential mortality among species 
and sizes of trees flooded at different depths when Lake Ocklawaha was created with 
impoundment of the Ocklawaha River, Florida. Lowe (1986) also related vegetation 
patterns to the hydrologic regime of a Florida lake, although he attributed most of the 
lake margin zonation to fire history. Robel (1962) reported changes in growth forms of 
sago pondweed {Potomogeton pectinatus) in response to altered hydrologic regime. 
Changes in pond cypress basal structure with flooding duration are described by Kurz 
and DeMaree (1934). Wetland vegetation zonation is also a response to water sources 
and the effects of physical hydrologic processes on the substrate (Bornette and Amoros 
1991). As indicated by these studies, even subtle alterations to a wetland' s hydrologic 
features may result in changes in the species composition, structural forms, and hence 
the wetland type. 

In an area undergoing succession, species occurrences are affected by light 
availability, substrate condition, proximity to seed or propagule source, and in a wetland, 



566 
the hydrologic regime. All of these factors are affected by the age and history of the site 
undergoing succession. Additional limitations as an area is colonized include a species' 
ability to use and sequester nutrients, compete for pollinators, and defend against 
herbivores. Species' plasticity to environmental change due to fires, freezes, wind, or 
flooding drives secondary succession; whether a community redevelops depends on the 
type and intensity of disturbance and the species's ability to grow and reproduce in spite 
of the altered conditions. Theoretically, any system that has reached a "climax state" is 
stable temporally and spatially only in a relative sense; succession is a cyclic process that 
occurs across the landscape at varying rates, creating a "moving mosaic" in response to 
disturbance events (White 1979). In a wetland the suite of species that can respond to 
the disturbance, perpetuating the succession cycle, is narrowed due to the physiological 
constraints of flooding. However, disturbance processes are part of the general 
phenomena of dynamics in the wetland community structure, and preservation of species 
in the landscape is dependent on preservation of the natural disturbance processes (White 
1979). 

The Suwannee River sill extended hydroperiods, decreased water depth 
variability, and increased water depths in approximately 18% of the Okefenokee Swamp 
(Chapter 3). Although alteration of the fire regime was intended with this structure, its 
greatest impact has been to extend high water depths during seasons less prone to 
wildfire occurrence (see Chapters 3 and 5). In 1990 the impounded region contained 
upland and wetland vegetation. Upland species were confined to the sand-based islands 
elevated above the river floodplain, however, and probably were not directly affected by 



567 
sill-induced changes to the surrounding environment, although the surrounding 

impounded conditions may have arrested fire movements off these islands (see Chapter 

5). Floodplain forests of pond cypress (Taxodium ascendens), blackgum (Nyssa sylvatica 

v. biflora), dahoon holly {Ilex cassine), loblolly bay (Gordonia lasianthus), sweetbay 

(Magnolia virginiana), and Carolina ash (Fraxinus caroliniana), areas of shrub and 

shrub-forest mix, and deep and shallow water prairies also occurred in the area that 

experienced increased flooding depth and duration, and decreased flooding variation. 

Previous studies by Glasser (1986, 1985), Best et al. (1984), Hamilton (1984, 1982), 

Deuver and Riopelle (1984a, 1984b, 1983), and Cypert (1973, 1972, 1961) examined 

responses of species in the swamp to fire and logging, and Trowell (1987) hypothesized 

that periodic freezes kill swamp vegetation which may later affect fire behavior. 

However, examination of the role hydrology plays in shaping the compositions and 

distributions of swamp vegetation communities has been limited (Deuver 1982, 1979). 

In order to predict changes in swamp vegetation that might occur as a result of sill 

manipulation, the hydrologic environment of current swamp vegetation species needed 

better description. This chapter discusses the following issues: 

1) What were the hydrologic environments during 1962-1995 at sites 
occupied by selected species during 1993-1994? 

2) Using these species-environment descriptions, what changes in species 
distributions may occur in response to alterations of the swamp hydrologic environment 
by manipulations of the Suwannee River sill? 



568 
Methods 

Vegetation Sampling 

During 1993 and 1994 vegetation was sampled in 5 regions of the swamp (Figure 
6-1). These areas were selected on the basis of accessibility and distance from the 
Suwannee River sill; as determined during initial reconnaissance, they included 
vegetation community types found throughout the swamp. Four of the regions were 
designated as prairies on 1964 USGS 1:24,000 topographic quadrangle maps (Chesser, 
Durdin, Floyd's, Sapling). The area bordered by the Suwannee River sill, Craven's 
Hammock, Billy's Lake, and the Pocket was designated the fifth sampling region (Sill 
Area). 

Each region was subdivided into 4 sections (Northwest, Northeast, Southeast, 
Southwest); within each of these, 4 transects of various lengths (30-120 m) were 
randomly located, traversing the topographic gradient nearest to the randomly located 
starting point and marked with PVC poles pushed into the surface peat (Figure 6-2). 
Many transects crossed peat-based island perimeters if that was the topographic gradient 
closest to the initial random location of the transect starting point. Other gradients 
crossed prairie perimeters or traversed the general topographic rise across the landscape. 
Structural diversity in the vegetation was apparent along the transect gradient and was 
used to delineate zones (or coenoclines) for sampling species composition associated 
with topographic and hydrologic gradients (Elton and Miller 1954). Descriptions of 



569 



03 

1 
I 

CD CO 

c75 q:^ 



a 



w 



c 

I 



CO 

"85 








o 
8 



S3 

i 

D 

o 

c 

— 

o 

OS 

ON 



O 
DO 

fi 

'C 

3 

"O 

•o 






o 

en 



c 
o 

I 

<u 

> 



e/i 

C 

o 

o 

o 



g 



570 



E 

o 



t! 



1 
g 

1 

i 

1 



B 

■3 

2 

oo 
# w 

■a 
o. 

2 

DC 

O 

a 

o 
— 

c 
o 



a 

- 

o 

« 

on 
c 

*S 

I 

! 

e 

WD 



**^*y^ 



III 







CQ 



u 
« M IS 
O ,S 2 

Ef § 

lis 




5 g.SPJi 



§ 
I 

s 

- 



X) 

8 



n 

_o 
"a. 



C 



o 

■*-• 
(/J 

B 

> 

o 

2? 

o 

o <u 

<D o 

G - 

u u 

ca _* 

0.0 
u c 

S a 

° ON 

§ rA 

&& 
.2 — 
'•5 bo 
o c 




571 
structural types recognized in the Okefenokee Swamp as vegetation zones are reported in 
Table 6-1. All transects ran from the deep to the shallow end of the water depth gradient. 
Along the transect PVC poles marked the transitions between the vegetation zones, 
indicated by vegetation structural changes or coenoclines. Each zone was further 
subdivided into 2-4 equal-length segments and marked with PVC poles. These sites 
provided replicate samples within the zones, and the transects provide sample replication 
within the area (Figure 6-2). 

During June-July 1993 and 1994 vegetation was sampled along all transects. 
Overstory data were collected in 1993, and understory sampling was conducted in 1994. 
For understory samples, a quadrat frame (0.5 m x 1 .0 m) was placed with the lower right 
corner at the PVC site marker within each zone, and the short axis parallel to the transect 
gradient (Figure 6-2). Species percent cover was estimated in 5% increments at 3 heights 
(< 0.3 m, 1.0 m, >1.0 m) above the ground surface; trace amounts were recorded as 1% 
cover. Cover totaled 100% at each height, and included estimates of open water, 
periphyton, and bare peat where appropriate. At the center of each quadrat, estimates of 
available photosynthetically active light (PAR, 0-199 /umol s'm" 2 ) at 0.3 m and 1.0 m 
were made with a Licor quantum sensor (LICOR, Inc., P.O. Box 4425, Lincoln, NE 
68504); a measurement was also made near the transect origin where no canopy cover 
occurred at the start and end of light sampling and used to standardize measurements to 
the total available light during the sampling effort. Light measurements were made only 
on cloud-free days. Measurements were also adjusted for daily variations in sun-horizon 
position (Astrolnfo, TUMASOFTware, Inc., Zephyr Services, 1900 Murray Avenue, 







572 




Table 6-1. Structural zone types recognized along sampled topographic/hydrologic 
gradients in Okefenokee Swamp. 






Structural Zone Type 


Abbreviation 


Description 




aquatic prairie 


aqupra 


deep water, floating vegetation with scattered 
herbaceous emergents, no overstory 




aquatic-herbaceous prairie 


aquher 


deep water, floating-emergent herbaceous 
vegetation mix with floating dominant, no 
overstory 






herbaceous prairie 


herpra 


moderate to shallow water, emergent 
herbaceous vegetation, no overstory 






aquatic prairie-trees 


aqutre 


deep water, floating and emergent herbaceous 
vegetation, moderately dense tree overstory 






aquatic prairie-shrubs 


aqushr 


deep water, floating and emergent herbaceous 
vegetation, moderately dense shrub overstory 






herbaceous prairie-trees 


hertre 


shallow water, emergent herbaceous 
vegetation, moderately dense tree overstory 






herbaceous prairie-trees- 
shrubs 


hertsh 


shallow water, emergent herbaceous 
vegetation, moderately dense tree and shrub 
mix overstory 






shrubs-herbaceous prairie 


shrher 


shallow water, emergent herbaceous 
vegetation, dense shrub overstory 






shrubs 


shrubs 


shallow to deep water, sparse herbaceous 
understory, dense shrub overstory 






shrubs-trees 


shrtre 


shallow to deep water, sparse herbaceous 
understory, dense shrubs with scattered trees 
in overstory 






trees-shrubs 


treshr 


shallow to deep water, sparse herbaceous 
understory, dense trees in overstory with 
scattered shrubs 






trees 


trees 


shallow to deep water, sparse herbaceous 
understory, dense tree overstory 












i 



573 
Pittsburgh, PA, 15217). A spherical densiometer (Forest Densiometers, 5733 Cornell 

Dr., Bartllesville, OK 74006) held at 1.5m above the peat surface was also used at each 
sample site to estimate overstory canopy cover. Water depths to the nearest 0.5 cm were 
recorded at the lower left, center, and upper right quadrat points; water depth at a staff 
installed at the transect origin was concurrently recorded. These staffs were calibrated to 
nearby (within 1000m) water level recorders by periodically measuring depths over time 
(every 3-4 months during 1992-1995) and noting changes in water surface elevation at 
the transects and recorders (see topographic survey discussion in Chapter 2); daily water 
surface elevations at the recorders could therefore be used to estimate daily water depths 
at sample sites along the transects when site water depths were not actually measured. 

All transects sampled for understory in 1994 were sampled for overstory 
composition in 1993. A 2 m x 2 m quadrat sharing a common lower right corner with the 
understory quadrat was measured at each overstory site (Figure 6-2). Percent cover of 
each woody overstory species estimated at 1 m, 2 m, and >2 m heights were recorded, 
and presence of stems <2.5 cm, 2.5-10.0 cm, and >10.0 cm dbh (diameter at breast 
height, 1.5 m above ground) within the quadrat was recorded by species. During 1994, 
transects randomly selected for seed bank composition analysis (see Chapter 7) were also 
sampled for additional shrub and tree composition. This information included a large 
number of most species along the transects, and provided estimates of species densities 
(number per m 2 ) along the transects. In each shrub sample quadrat along a transect, 
stems of each shrub species were counted within a 0.5 m x 2 m quadrat placed with the 
site marker at the lower right corner, and the quadrat's short side parallel to the transect 



574 
gradient (Figure 6-2). Several quadrat sizes and orientations were initially sampled; this 
dimension and placement provided the highest densities and species richness for the area 
sampled. A belt extending the transect' s length and 5 m out from either side and 2 m 
beyond the last sample site, was used to describe the transect's tree composition (Figure 
6-2). All trees >1.0 m tall within the belt were identified, counted, and a dbh 
measurement recorded; shrub species were not recorded in this sample. Water depths 
were estimated for the shrub quadrats using relationships established for the understory 
samples; water depths for the understory quadrats were assumed representative of those 
in the 0.5 m x 2.0 m shrub quadrats. Estimates of the water depths for the trees recorded 
along the belt transects were made by structural zone; water depths measured at the 
understory quadrats were averaged across the zone, and this value represented the water 
depth for species encountered in the zone. Long-term water level data for structural 
zones in the belt transects were estimated from recorder data as with understory quadrats. 
Preparation of Hvdrologic Data 

Although water depth may limit distributions of some species, duration of 
inundation and variability in water levels may also affect species' occurrences. Daily 
water level data recorded and estimated at sites in the areas of the vegetation transect 
sampling were available for 1941-1995 (see chapter 2). These data were used in 
assessment of the hydrology model discussed in Chapter 3. Extension of these data to 
the vegetation transects for calculation of water depths, water depth variability, and 
flooding duration at each site during 1941-1995 were made based on the elevation 



575 
relationships among sampled sites, staffs in place at each transect, and the nearest 
(within 1000m) water level recorders. These transect-recorder pairs are listed in Table 6- 
2. Daily water depths estimated for each transect site were summarized in LOTUS 123 
spreadsheets into several variables. Average daily water depth during sill gate closure 
(1962-June 1995) was calculated at each site; depths during this interval were not 
different statistically from those summarized by decades during the same period (Table 
2-20). Average water depths at each sampled site were also calculated for growing 
(March-October) and non-growing (November-February) seasons during 1962-June 1995. 

Duration of inundation with sill gate closure (1962-June 1995) was also 
calculated for each quadrat and for several inundation depths (depth classes, denoted DC 
in figures and tables), providing an indication of whether a species was found where peat 
was usually inundated, and also a description of the inundation depth. Inundation depth 
classes were defined by relationships of general plant height to water depth (Table 6-3). 
Reliability of water depths measured below the peat surface were uncertain, so minimum 
estimated depths were summarized as < m. This indicated soils that were dry, moist, or 
possibly saturated, but the surface was not inundated. Water depths > m and < 0.30 m 
("shallow" water depth) indicated submergence of at least the bases of herbs and shrubs, 
but not necessarily trees. Depths > 0.30 m ("deep" water depth) indicated submergence 
of most tree bases. These depths were further subdivided to examine differences in 
peaks of species occurrence and average daily water depths. Estimated water depths 
>0.00 m-0.05 m represented peat that was inundated, but plants were generally not 
submerged. Smaller stature herbs were submerged by water depths 0.05 m-0. 15 m; 



576 



Table 6-2. Recorders and nearest survey benchmarks used to estimate water surface 
elevations at vegetation transects during 1960-1995. 



Area 


Recorder 


Survey 
Benchmark 8 


Transect b 


Chesser Prairie 


Seagrove Lake 


17 


8,9,14 




Seagrove lake 


18 


10,15 




Chesser Prairie 


1 


1,3,4,7,12,13 




Chesser Prairie 


2 


2,5,6,11,16 


Durdin Prairie 


Kingfisher Landing 


31 


18 




Durdin Prairie 


32 


17,19,22,23,25 




Durdin Prairie 


33 


20,21,24,26,29 




Durdin Prairie 


34 


27 




Durdin Prairie 


54 


28,30,31,32 


Sapling Prairie 


Sapling Prairie 


22 


74, 75, 76, 77 




Sapling Prairie 


23 


64, 65, 66 




Sapling Prairie 


24 


69,70,71,72,73 




Sapling Prairie 


25 


67,68 




Sapling Prairie 


53 


78,79 


Floyd's Prairie 


Floyd's Prairie 


44 


34, 46, 47 




Floyd's Prairie 


45 


37, 42, 43, 45 




Floyd's Prairie 


46 


38,39 




Floyd's Prairie 


47 


41,48 




Floyd's Prairie 


48 


33, 35, 36, 40, 44 


Sill Area 


Suwannee River 


57 


49,80 




Suwannee River 


58 


50,63 




Suwannee River 


59 


58 




Suwannee River 


63 


51,53,54,60 



577 



Area 


Recorder 


Survey 
Benchmark" 


Transect b 




Sill (Brown Trail) 


60 


57 




Sill (Brown Trail) 


61 


56 




Sill (Brown Trail) 


64 


52,55,59,61,62 



a Survey benchmarks were within 1000m of transect locations. 
b Transect locations are listed in Appendix C. 






578 



Table 6-3. Inundation depth classes defined for analysis of species occurrence in 
hydrologic environments. 



Depth Class 
(DC) 


General Inundation 
Description 


Water Depth 
Range (m) 


Extent of Plant 
Submergence 


DC1 


no inundation 


depth < 0.00 


no inundation 


DC2 


shallow 


0.00 < depth < 0.05 


inundated peat; 

small plants not 

submerged 


DC3 


shallow 


0.05 < depth < 0.15 


small herbs 
submerged 


DC4 


shallow 


0.15 < depth < 0.30 


large herbs and 

bases of shrubs 

submerged 


DC5 


deep 


0.30 < depth < 0.60 


tree bases 
submerged 


DC6 


deep 


0.60 < depth < 1.00 


tree bases 

submerged; 

common in sill area 


DC7 


deep 


depth > 1.00 


tree bases 

submerged; 

common in sill area 



579 
larger herbs and the bases of most shrubs were submerged at water depths 0. 15 m-0.30 
m. Tree bases were generally submerged when water depths exceeded 0.30 m. Further 
subdivisions of 0.60 m-1.0 m and >1.0 m permitted examination of species occurrences 
in extreme water depths common in the sill-affected area (Table 6-3). The number and 
proportion of days during 1962- June 1995 that a quadrat was in each of these 7 water 
depth categories was totaled. Percentages were combined to calculate proportions for 
combined depth classes, particularly depth < m (no inundation), < depth < 0.30 m 
(shallow inundation), and depth > 0.30 m (deep inundation). 
Analysis of Vegetation Data 

Percent cover estimates provided information about species occurrence from two 
perspectives. At the landscape level, environments of sampled quadrats represented the 
suite of hydrologic conditions available throughout the swamp, regardless of species 
occurrence. Species percent cover estimates for these samples were logit-transformed 
(y=//i[p/l-p], where p = species percent cover) to normalize skewed distributions due 
infrequent species occurrence. Site descriptions where species were present represent 
the environment on a smaller, local scale, without consideration of the swamp- wide 
environment (which includes areas where species were absent). Therefore, datasets were 
re-sampled to include only quadrats where species occurred, so that species-environment 
relationships could be examined. 

Comparisons of conditions where species occurred with conditions where species 
were absent suggested local and landscape-level differences in hydrologic environments. 



580 
Examination of occupied sites refined the site descriptions beyond features that 
determined species presence or absence, to indicate conditions most favorable to species' 
abundances. Mest (average water depth) and Wilcoxon rank-sum (percent of interval in 
each depth class) procedures were used to identify differences in species abundance 
among hydrologic conditions. 

Statistically significant relationships among species occurrences and 
environmental variables were identified using a mixture experiment format. In mixture 
experiments, frequently used in agricultural research to analyze suitability of component 
blends (such as proportions of juices in fruit juice blends), the measured characteristic 
(e.g., juice preference) is assumed to be dependent on the relative proportions in each of 
the mixture ingredients (Cornell and Harrison 1997). Location of a point (the juice blend 
"suitability score") in factor space can be described by a multiple regression model. The 
^-dimensional model can be visualized in an n-dimensional plot or surface, to illustrate 
interaction affects among components and significance of components in affecting the 
measured characteristic (suitability score) that are identified in development of the best 
regression model. These surfaces (models) can be statistically compared with F-tests to 
determine similarities of measured characteristics among different mixtures. 

For analysis of plant species association with hydrologic environments, the 
species abundance (representing the suitability of the hydrologic environment of a 
sampled quadrat), could be described by the proportion of time a site spent in each water 
depth condition (no inundation, shallow, deep), interactions of time spent at these depths, 
and the covariate effects of light availability and transect. The general model form 



581 
(Cornell and Harrison 1997) of the 3 -dimensional model is: 

response = p,x, + p 2 x 2 + p 3 x 3 + p 12 x,x 2 + p 13 x,x 3 + P 23 x 2 x 3 + (P 123 x,x 2 x 3 }+ e 

where response is the species' abundance, P's are estimated values describing the 
relationships among the components (species occurrence and hydrologic conditions) in 
the experimental data, and x„ represents the duration of flooding in each depth. The 3- 
dimensional model was chosen because the hydrologic environment could be described 
by 3 proportions describing the duration and of inundation and totaling 100% of the 
sample interval (% time with no inundation+% time with shallow inundation+% time 
with deep inundation=100% time). The term (P, 23 x,x 2 x 3 } was replaced in this analysis 
with the covariates (light availability and transect) and their interactions. The 3- 
dimensional models were developed and significance of parameters assessed using the 
SAS version 6. 12 Proc GLM procedure (SAS Institute, Inc., Cary, NC 27513). Model 
reduction was based on Mallow's C p (Myers 1990) and effects of forward and backward 
addition of components to changes in Type EH sums of squares. Models were similarly 
assessed for entire data sets (all sampled quadrats) and reduced data sets (quadrats only 
where species present). 

Species models that indicated a significant relationship between the species 
abundance and no inundation to shallow inundation conditions (0-0.30 m water depth) 
were modeled again to determine if abundance differed among inundation depths of 0- 
0.05 m, 0.05-0.15 m, and 0.15-0.30 m. The models were developed and significance of 



582 
parameters assessed using the SAS Proc GLM procedure as discussed above. Models 
were similarly assessed for entire data sets (all quadrat data regardless of species 
presence) and reduced data sets (quadrat data only where species were present). 

In addition to species richness, Shannon- Weiner diversity indices (Kent and 
Coker 1992) were calculated for all understory, shrub, and tree plots to assess differences 
in species diversities among hydrologic environments and sample regions (Chesser 
Prairie, Durdin Prairie, Floyd's Prairie, Sapling Prairie, sill area). Diversity measures 
were modeled as described above. 

Observed and model-predicted density or cover estimates were diagramed in 3- 
dimensional plots to visualize the shape of the modeled relationships (Figure 6-3). 
Similarities among species occurrences and hydrologic conditions were more easily 
visualized when illustrated in this manner; species could be grouped based on common 
plot shape, representing species with similar relationships between abundance and the 
modeled hydrologic parameters. Plots were constructed with SigmaPlot software 
(version 2.01, Jandel Corporation, San Rafael, CA 94912) using observed and model- 
predicted abundances calculated with SAS-Proc GLM procedures described above, and 
blindly (without knowledge of plot species identification) clustered by common plot 
shapes to determine if species' groups or associations might exist. 

Changes in the swamp hydrologic environment predicted from with-sill and 
without-sill hydrology models (Table 3-6) were compared with diagramed and modeled 
species-environment relationships to ascertain vegetation changes that might occur with 






583 









GRAPH 1 



Water depths shallow or 
no inundatin, but never deep. 



Water depths frequently 

shallow; seldom no inundation 

or deep water. 




Frequently no inundation; 
seldom inundated. 



Frequently deep water; 
seldom shallow or no inundation. 



Note: % of time in deep water is represented at XY origin, and is calculated as 
100% - (% time with no inundation + % time in shallow water). 



Figure 6-3. Example interpretation of axes (Graph 1) and curvatures (Graph 2) on 3- 
dimensional plots of model-predicted abundances of species with flooding depth and 
duration. 



584 



GRAPH 2 



Curvature in this region 

indicates significant interaction 

between no inundation 

and shallow inundation. 



Curvature in this 

region indicates 

significant interaction 

between shallow and 

deep inundation. 




100 e < 

80 ■& 



' nu n<*stio n 



Curvature in this region 

indicates significant interaction 

between no inundation 

and deep inundation. 









Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-3-contimieH 



585 
sill removal. Changes in hydrologic environments that could lead to species changes 
were summarized by swamp region. 

Results 

Species' Environments 

Most species occurred at similar average daily water depths (Figure 6-4). Figure 
6-5 illustrates the gradients from deep to shallow and constant to variable water depths 
occurring along the sampled transects. Figure 6-6 indicates substantial overlap in 
species' abundances (for a composite of all samples) across the exposure duration 
gradient; however, differences among species and species groups emerge when the 
duration and degree of inundation are isolated. Descriptors of the percent of time spent 
in each depth class are listed in Table 6-4 for 49 species occurring in at least 3 of the 944 
understory plots, 489 shrub plots, or 166 tree belt samples. Hydrologic conditions where 
species were absent are listed in Table 6-5. These comparisons indicate that most of the 
sampled species occur under specific conditions of light availability, hydroperiod, and 
inundation depth in the swamp. Frequency quartiles of flooding durations in each water 
depth range, for locations where species abundances were greatest (90-100% of the 
maximum cover or density), are described in Figure 6-7. For each water depth range or 
depth class, the most frequent (mode) duration of inundation, the maximum flooding 
duration (range) where the species occurred, and the maximum flooding duration (range) 
of all sampled quadrats are also indicated. These summaries suggested that species 



9 

> 
O 

o 



100 - 
80 - 
60 
40 
20 
- 



100 - 
80 
60 - 
40 - 
20 - 


100 - 
80 
60 - 
40 - 
20 - 
- 

100 - 



Carex walteriana 



Vi*»' ' 



— A 




Peltandra saggiti folia 



«K \ 



Eleocharis vivipara 



- • 



.-. \m-Yh 



tm — — 



586 




Average Daily Water Depth (m) 



Figure 6-4. Average daily water depths (1962-1995) for species recorded at Okefenokee 
Swamp sample sites during 1993-1994. 



0) 

> 
o 
o 



i- 
0) 

> 
o 

O 



> 
O 

o 



100 
80 
60 
40 
20 


100 
80 
60 
40 
20 


100 
80 
60 
40 
20 


100 
80 
60 
40 
20 


100 
80 
60 
40 
20 


100 

80 

60 

40 

20 



100 

80 

60 

40 - 

20 
- 



Utricularia spp. 




•\ - 



... . ..... • "V« • 9 

Woodwardia virgmica • »# 




•• •• • 



Xyris spp. 






Lacnanthes caroliniana 



• • • 

• •• » • 

• •• • # • 

» v irti ■!■■ * i 



31 



J_ 



I I 



Eriocaulon lineare 



MM* 



I I 



»• — > 



Nuphar luteum 



.-«—<. 



• • • 



«■■ 



li lt 



/tea virginica 



•*• • • 



i i i i i i i i 1 1 r 



587 




■fa* m\ aj 

'' ii r 



-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 

Average Daily Water Depth (m) 



Figure 6-4-cnntinned 






E 

5 



8 
6 
4 
2 



CO 

c 
<p 





o 


40 


N 


30 


E 






20 


a 




e 

l. 


10 


< 




« 





« 




a 


20 


m 





15 

10 

5 




3 
2 
1 


25 

20 

15 

10 

5 


3 



1 - 



CO 

a> 

< 



00 




25 

20 

15 

10 

5 





Smilax walteri 




I I 



Gordon/a lasianthus 



— • . XaliAi 



//ex cassine 



Magnolia virgin iana 



•• • *««dktfjh&rffe 



>• • • 



i r 



Pin us spp. 






I 



Persea palustris 



M • M 



I ' i I I I I" 



»•• • • 



Nyssa sylvatica 



I I 1 l' | 



588 



I 1 



I I 



i i i i 



i i i r~ 

-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 

Average Daily Water Depth (m) 



Figure 6-4-continued 



E 



CO 

c 

0) 

Q 



8 

6 

4 

2 



8 

6 

4 

2 


15 
12 - 

9 - 

6 

3 


80 

60 - 



E 

J 40- 

» 20- 

a> 

r\ 



15 - 

10 - 

5- 



-i 

40 



E 

i 

to 

c 
a> 
a 



30 
20 - 
10 - 


50 
40 - 
30 - 
20 
10 





Cephalanthus occidentalis 



589 



MLt. .— . 



Clettira alnifolia 






X 



X 



Cyrilla racemiflora 




Lyonia lucida 



• • 



s. a.- 




• • 



Leucothoe racemosa 



. _Cua 



Pierua phillyreifolia 






i i i r 



i ' 



Smilax laurrfolia 



T 



■HS*- 



.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 

Average Daily Water Depth (m) 



0.8 1.0 



Figure 6-4-continued 



20 



.j g _ Fraxinus caroliniana 



10 - 



5 - 



- 
200 



E 150 - 

■ 
■ 
2 100 - 



% 50- 



- 



590 





6 - 


Acer rubrvm 


E 






-— . 






a 


4 - 




9 






u 






< 








2 — 




(0 






(0 






«8 






00 


- 





• •« 



30 - 


Nyasa ogeechee 










20 - 








• 




10 - 












- 






• 


• 


mm • • • 










I I I 


I I 


I 




I I I I I I 



Taxodium ascendens 



• • 






- • •« 



i n i i i i i 1 1 — i 1 r 

-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 



Average Daily Water Depth (m) 



Figure 6-4-continued. 



591 




o 
eg 



_ oo 



_ <o 



c 

s 

o 

(0 

c 

05 



0) 
LLQOC0 



eg 



_ o 



- 00 



CO 



(1) 0) 



i — r 



o in o 
io * * 



i i i — i — r 



CM 



io o in o m 
co co cm eg t- 



o io o 

r- O O 



If) O 
O t- 



~i i i i — i — r 

U) O IX) O U) o 
T- CM CM CO CO •<* 



ooooooooooooooooooo 

Ill 

(uu) ifldaa ja^BM Ahbq aBejeAy 



c 
o 



On 
CTs 



o 
o 

00 

c 
'C 

3 
T3 

o. 
£ 

en 

o 

<u 
_<: 

o 

c 

o 

o 

-o 



t/5 



c 


u 


O 


S 


4—' 


C 


(0 


2 


o 


«-^ 


CL 


c 
o 




ccj 


CO 




<D 


6 

> 


Q. 


01] 


E 


c 
o 


CO 


CO 


CO 


Vi 



<u 

"O 
u. 
<D 

! 

1 

CO 

i> 

i 

VO 



592 




J9A0Q o/ uiniu;xe|/\| 



593 

Table 6-4. Hydrologic environments during 1962-1995 of species occurring in 
vegetation sample plots during 1993-1994. Water depth conditions (DC) are described in 
the table footnote. 



Parameter 


Carex 
walteriana 


Nymphaea 
odorata 


Xyris spp. 


Utricularia spp. 


Sample Size 


437 


361 


248 


244 


Water Depth * + SD (m) 


0.17 + 0.14 


0.24 + 0.13 


0.17 + 0.12 


0.29 + 0.14 


Minimum Water Depth (m) 


-0.40 


-0.28 


-0.40 


-0.17 


Maximum Water Depth (m) 


0.55 


0.62 


0.44 


0.95 


R %TimeinDCl(SD) * 


17.7(20.8) 


11.4(16.0) 


13.0(18.9) 


8.9(12.2) 


x % Time in DC2 (SD) 


6.1 (4.3) 


4.1 (3.7) 


6.6 (5.4) 


3.4(2.9) 


x % Time in DC3 (SD) 


18.7(12.5) 


13.7(12.2) 


25.1 (16.9) 


11.2(9.5) 


8 %TimeinDC4(SD) 


30.5 (14.2) 


30.2 (15.7) 


36.0(18.6) 


29.1(14.4) 


x % Time in DC5 (SD) 


24.8 (20.2) 


37.1 (24.2) 


18.3(20.3) 


41.8(21.6) 


x % Time in DC6 (SD) 


2.1 (3.9) 


3.3(6.3) 


1.0(1.8) 


4.8 (7.7) 


x %TimeinDC7(SD) 


0.1(0.8) 


0.2(1.1) 


0.0(0.1) 


0.9 (5.0) 


Mode of %Time in DC1 














Mode of %Time in DC2 














Mode of %Time in DC3 


27.5 





49.1 





Mode of %Time in DC4 


26.8 


27 


32.1 


26.8 


Mode of %Time in DC5 


1.9 


0.2 


1.9 


31.1 


Mode of %Time in DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC1 


0,95 


0, 88.9 


0,95 


0, 78.6 


Minimum, Maximum % Time in DC2 


0, 27.5 


0,24 


0, 35.7 


0, 17.1 


Minimum, Maximum % Time in DC3 


0,65.1 


0, 59.6 


0,64.4 


0, 57.8 


Minimum, Maximum % Time in DC4 


1,76.2 


1.2,79.9 


0.8, 79.9 


0.3, 74.4 


Minimum, Maximum % Time in DCS 


0, 96.8 


0.1,98.2 


0, 98.2 


0.1,98.3 


Minimum, Maximum % Time in DC6 


0, 36.7 


0,53.1 


0,11.7 


0,53.1 


Minimum, Maximum % Time in DC7 


0, 12.0 


0, 12.0 


0,0.5 


0, 44.3 


Overstory % Cover, x + SD 


29.3 + 36.8 


13.2 + 27.1 


25.5 + 35.1 


20.8 + 34.1 


% Low Level Light Available, x + SD 


47.6 + 35.0 


68.6 + 34.3 


56.5 + 35.5 


60.3 + 38.8 


% High Level Light Available, x + SD C 


68.5 + 33.3 


82.7 + 25.6 


71.0 + 33.0 


74.6 + 30.7 



Table 6-4-continued 



594 















Eriocaulon 


Rhynchospora 


Eleocharis 


Panicum 


Parameter 


lineare 


inundata 


baldwinii/ 
vivipara 


hemitomon 


Sample Size 


44 


35 


157 


228 


Water Depth x + SD (m) 


0.2+0.1 


0.17 + 0.13 


0.16 + 0.20 


0.19 + 0.15 


Minimum Water Depth (m) 


-0.10 


-0.17 


-0.31 


-0.30 


Maximum Water Depth (m) 


0.40 


0.37 


0.67 


0.55 . 


x % Time in DC 1 (SD) * 


5.3 (9.0) 


13.6(21.8) 


30.9(21.5) 


15.8(20.0) 


x % Time in DC2 (SD) 


6.9 (4.7) 


5.8(7.2) 


6.3 (5.0) 


5.4 (4.8) 


x % Time in DC3 (SD) 


39.3(19.7) 


19.8(17.6) 


16.5(13.3) 


18.1 (14.9) 


x % Time in DC4 (SD) 


38.5(19.1) 


36.4(21.9) 


19.8(15.6) 


29.9(16.4) 


x % Time in DC5 (SD) 


9.9(21.4) 


24.1(25.1) 


13.8(11.5) 


27.5 (24.0) 


x % Time in DC6 (SD) 


0.0(0.1) 


0.2 (0.6) 


7.1 (8.6) 


2.8(5.3) 


x % Time in DC7 (SD) 


0.0 (0.0) 


0(0) 


5.5 (8.3) 


0.5(2.3) 


Mode of %TimeinDCl 


6.7 











Mode of %Time in DC2 


10.1 





2.8 





Mode of %Trme in DC3 


49.1 


0.2 


5.5 





Mode of %Time in DC4 


32.1 


28.7 


12.3 


28.7 


Mode of %Time in DCS 


1.9 


0.1 


12.3 


0.8 


Mode of %Time m DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC1 


0, 59.2 


0, 78 6 


0, 88.9 


0,90.4 


Minimum, Maximum % Time in DC2 


0, 17.3 


0,35.7 


0, 36.5 


0, 35.7 


Minimum, Maximum % Time in DC3 


0,65.1 


0, 59.6 


0.4, 57.6 


0, 59.6 


Minimum, Maximum % Time in DC4 


1.2,76.8 


1.6,79.9 


2.0, 76.4 


1.2,79.9 


Minimum, Maximum % Tune in DC 5 


0.1,98.2 


0, 89.6 


0,66.3 


0, 98.2 


Minimum, Maximum % Time in DC6 


0,0.5 


0,2.9 


0,29.1 


0, 36.7 


Minimum, Maximum % Time in DC7 


0,0 


0,0 


0, 34.0 


0, 18.2 


Overstory % Cover, x + SD 


10.9+16.9 


10.9+17.5 


21.4 + 32.7 


16.7 + 28.5 


% Low Level Light Available, x + SD 


66.0 + 33.3 


71.5 + 30.1 


49.4 + 35.0 


61.3 + 34.9 


% High Level Light Available, x + SD c 


88.0+18.1 


87.4+19.7 


70.8 + 29.4 


79.7 + 25.9 



Table 6-4-continued 



595 















Lacnanthes 


Nuphar 


Woodwardia 


Peltandra 


Parameter 


caroliniana 


luteum 


virginica 


virginica 


Sample Size 


266 


68 


197 


356 


Water Depth x + SD (m) 


0.10 + 0.16 


0.26 + 0.21 


0.12+0.19 


0.15 + 0.14 


Minimum Water Depth (m) 


-0.45 


-0.31 


-0.29 


-0.45 


Maximum Water Depth (m) 


0.53 


0.81 


1.76 


0.53 


x % Time in DC 1 (SD)* 


28.5 (25.4) 


23.0(19.8) 


25.0 (24.2) 


19.8(22.1) 


S % Time in DC2 (SD) 


7.5 (4.7) 


3.5 (2.9) 


8.5(5.5) 


6.9 (4.7) 


x % Time in DC3 (SD) 


21.8(15.0) 


11.2(11.8) 


24.9(15.8) 


21.7(14.4) 


x % Time in DC4 (SD) 


25.0(16.2) 


22.5 (22.8) 


26.0(15.6) 


30.2(15.0) 


x % Time in DC5 (SD) 


13.4(15.4) 


20.6(18.8) 


13.3(17.9) 


19.8(19.1) 


x % Time in DC6 (SD) 


2.4 (5.2) 


11.0(9.0) 


1.7(4.7) 


1.5(3.3) 


x % Time in DC7 (SD) 


1.3(4.5) 


8.2 (9.5) 


0.6 (6.5) 


0(0.1) 


Mode of %TimeinDCl 


6.7 





6.71 





Mode of %Time in DC2 


10.1 


2.2 


10.1 





Mode of %Timc in DC3 


49.1 


5.5 


49.1 


49.1 


Mode of %Time m DC4 


32.1 


12.3 


32.1 


32.1 


Mode of %Time in DC5 


0.2 


19.0 





0.1 


Mode of %Time in DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC 1 


0, 96.6 


0, 78.2 


0, 96.8 


0,96.6 


Minimum, Maximum % Time in DC2 


0, 36.5 


0,13.1 


0, 35.7 


0, 35.7 


Minimum, Maximum % Time in DC3 


0.4,65.1 


0, 50.6 


0,65.1 


0,65.1 


Minimum, Maximum % Time in DC4 


0.5, 76.4 


0.3, 76.8 


0.1,80.5 


0.2, 79.9 


Minimum, Maximum % Time in DC5 


0,66.3 


1.2,98.3 


0, 98.2 


0, 98.2 


Minimum, Maximum % Time in DC6 


0, 32.0 


0,29.1 


0, 32.0 


0, 32.0 


Minimum, Maximum % Time in DC7 


0, 26.9 


0,35.7 


0,90.5 


0,0.9 


Overstory % Cover, x + SD 


28.6 + 35.6 


9.6 + 26.5 


39.6 + 38.1 


32.8 + 36.6 


% Low Level Light Available, x +SD 


45.7 + 35.3 


48.9 + 35.8 


33.2 + 33.6 


44.8 + 35.5 


% High Level Light Available, x +SD° 


69.0 + 33.3 


77.7 + 21.1 


59.4 + 37.0 


64.5 + 35.0 






Table 6-4-continued 






596 















Sphagnum 


Andropogon 


Dutichium 


Orontium 


Parameter 


spp. 


virginica 


arendinacium 


aquaticum 


Sample Size 


308 


61 


162 


133 


Water Depth x + SD (m) 


0.18 + 0.26 


0.17 + 0.06 


0.18 + 0.24 


0.26 + 0.12 


Minimum Water Depth (m) 


-0.22 


-0.04 


-0.21 


-0.17 


Maximum Water Depth (m) 


1.76 


0.35 


1.66 


0.62 


x % Time in DC 1 (SD)* 


20.4(21.9) 


5.8(5.8) 


20.6 (20.3) 


10.2(12.9) 


x % Time in DC2 (SD) 


7.7(5.7) 


6.5(4.1) 


6.8(4.1) 


4.2 (4.6) 


x % Time in DC3 (SD) 


24.4(16.8) 


34.9 (16.4) 


21.7(13.7) 


12.8(12.2) 


x % Time in DC4 (SD) 


28.6(18.9) 


42.2 (16.6) 


29.8(17.2) 


28.0(14.1) 


x % Time in DC5 (SD) 


13.4(17.4) 


10.2(14.0) 


15.7(14.9) 


42.3 (25.5) 


x % Time in DC6 (SD) 


2.5 (6.2) 


0.4(1.0) 


2.7(5.7) 


2.6 (5.6) 


x % Time in DC7 (SD) 


3 1(13.1) 


0(0) 


2.8(12.7) 


0.1(0.2) 


Mode of %Time in DC 1 





6.7 








Mode of %Time in DC2 





10.1 








Mode of %Time in DC3 


49.1 


49.1 


20.1 





Mode of %Time in DC4 


32.1 


32.1 


26.1 


26.6 


Mode of %Time in DC5 


0.1 


1.9 


0.8 


35.3 


Mode of %Time in DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC1 


0.88.1 


0,32.5 


0, 82.3 


0, 78.6 


Minimum, Maximum % Time in DC2 


0, 36.5 


0, 18.7 


0, 19.7 


0, 35.7 


Minimum, Maximum % Time m DC3 


0,65.1 


1.2,58.2 


0.4, 56.2 


0, 59.6 


Minimum, Maximum % Tune in DC4 


0.6, 80.5 


12.2,76.8 


0.7, 76.8 


1.5,79.9 


Minimum, Maximum % Time in DC5 


0, 98.2 


0, 74.0 


0,61.3 


0,98 


Minimum, Maximum % Time in DC6 


0, 32.0 


0,4.3 


0, 29.5 


0,53.1 


Minimum, Maximum % Tune in DC7 


0,90.5 


0,0.2 


0, 88.2 


0,2.1 


Overstory % Cover, x + SD 


24.7 + 34.0 


8.2 + 19.5 


26.8 + 35.7 


12.0 + 25.3 


% Low Level Light Available, x +SD b 


49.8 + 36.3 


68.1+28.7 


49.6 + 3S.2 


67.2 + 37.3 


% High Level Light Available,* +SD° 


71.8 + 32.6 


88 4 + 18.0 


72.1+31.1 


84.5 + 25.5 



Table 6-4--continued. 



597 















Saggetaria 


Triadenum 


Sarracenia 


Sarracenia 


Parameter 


graminea 


virginicum 


flava 


psittacenia 


Sample Size 


66 


42 


61 


15 


Water Depth x + SD (m) 


0.15 + 0.10 


0.14 + 0.08 


0.13+0.10 


0.12 + 0.05 


Minimum Water Depth (m) 


-0.17 


-0.17 


-0.17 


-0.06 


Maximum Water Depth (m) 


0.35 


0.29 


0.41 


0.16 


8 % Time in DC 1 (SD)* 


14.9(18.8) 


7.9(16.4) 


10.1(18.1) 


8.4(13.5) 


x % Time in DC2 (SD) 


6.9 (4.6) 


6.7(4.5) 


5.8 (0.5) 


8.4(3.1) 


x % Time in DC3 (SD) 


26.8(16.2) 


36.6(17.0) 


38.8(18.9) 


47.8 (9.9) 


x % Time in DC4 (SD) 


35.3(16.9) 


43.0(20.2) 


33.9(18.1) 


33.5 (9.7) 


x % Time in DC5 (SD) 


15.4(16.1) 


5.6(7.7) 


8.5(19.4) 


1.9(1.4) 


x % Time in DC6 (SD) 


0.6(1.0) 


0.1(0.6) 


(0.2) 


0(0) 


x % Time in DC7 (SD) 


0(0) 


0(0) 


0(0) 


0(0) 


Mode of %TimeinDCl 





0.1 





1.9 


Mode of %Time in DC2 


5.7 


0.6 





6.4 


Mode of %Time in DC3 


15.9 


26.8 


49.1 


18.2 


Mode of %Time in DC4 


30.3 


16.5 


32.1 


15.2 


Mode of %Time in DC5 


1.2 


2.3 


0.2 


1.2 


Mode of %Time in DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC1 


0, 78.6 


0, 78.6 


0, 76.6 


1.9,56.6 


Minimum, Maximum % Time in DC2 


0,24 8 


0.3,17.1 


0, 24.0 


4 8,17.1 


Minimum, Maximum % Time in DC3 


1.2,57.6 


5.5, 57.8 


0,65.1 


18.2, 58.3 


Minimum, Maximum % Time in DC4 


2.5, 76.4 


2.5, 76.8 


2.8, 76.0 


15.2,49.5 


Minimum, Maximum % Time in DC5 


0, 74.0 


0.1,41.5 


0.1,96.8 


0.4, 5.5 


Minimum, Maximum % Time in DC6 


0,3.8 


0,3.5 


0,1.4 


0,0 


Minimum, Maximum % Time in DC7 


0,0.1 


0,0 


0,0 


0,0 


Overstory % Cover, x + SD 


15.3 + 26.5 


3.7+10.6 


12.6 + 16.7 


14.3 + 22.6 


% Low Level Light Available,? +SD b 


62.0 + 30.6 


70.6 + 27.3 


54.6 + 31.4 


48.8 + 27.5 


% High Level Light Available, x +SD° 


80.8 + 23.7 


89.9 + 17.1 


80.5 + 24.9 


83.1+28.1 





Table 6-4~continued. 








598 






Eleocharis 


Iris 


Decodon 


Rhynchospora 




Parameter 


robbinsii 


virginiana 


verticillatus 


chalerocephala/ 
wrightiana 


Sample Size 


91 


43 


26 


44 




Water Depth x + SD (m) 


0.30 + 0.07 


0.18 + 0.08 


0.19 + 0.14 


0.10 + 0.14 




Minimum Water Depth (m) 


0.10 


-0.09 


-0.11 


-0.17 




Maximum Water Depth (m) 


0.62 


0.33 


0.50 


0.40 




x % Time in DC 1 (SD)* 


6.3 (4.5) 


11.6(10.8) 


10.9(16.5) 


20.1 (27.8) 




x % Time in DC2 (SD) 


3.0(1.6) 


7.8(5.2) 


6.6 (7.6) 


7.7(5.7) 




x % Time in DC3 (SD) 


10.1 (4.4) 


26.7(14.2) 


25.1 (18.7) 


30.3(18.7) 




x % Time in DC4 (SD) 


30.7(7.2) 


33.7(12.2) 


37.6 (23.6) 


33.6(23.0) 




x % Time in DCS (SD) 


44.5 (9.5) 


18.8(15.6) 


15.31 (17.0) 


8.0(16.8) 




x % Time in DC6 (SD) 


5.3 (6.0) 


1.4(1.6) 


3.9(8.6) 


0.11(0.5) 




x % Time in DC7 (SD) 


0.15(0.3) 


0(0.1) 


0.5(1.3) 


0(0) 




Mode of %Time in DC 1 


2.6 


3.7 










Mode of %Time m DC2 


2.1 


2.2 


0.3 


8.3 




Mode of %Time in DC3 


6.8 


7.8 


3.0 


11.2 




Mode of %Time in DC4 


39.3 


31.1 


2.0 


3.7 




Mode of %Time m DC5 


33.9 


2.3 


5.1 


0.2 




Mode of %Time in DC6 


2.7 













Mode of %Time in DC7 
















Minimum, Maximum % Time in DC1 


0. 27.9 


0,64.9 


2.4, 48.7 


0, 76.6 




Minimum, Maximum % Time in DC2 


0,11.2 


0, 24.0 


1.6, 16.3 


0, 27.5 




Minimum, Maximum % Time in DC3 


0.5,23.7 


2.3,52.9 


5.9, 25.3 


0,61.7 




Minimum, Maximum % Time in DC4 


3.1,59.5 


7.1,73.9 


15.2,24.7 


3.2, 76.8 




Minimum, Maximum % Time in DC5 


12.4,60.8 


0.4, 49.7 


3.5, 57.3 


0, 92.3 




Minimum, Maximum % Time in DC6 


0,53.1 


0,5.3 


0.2, 7.7 


0,3.5 




Minimum, Maximum % Tune in DC7 


0,2.1 


0,0.4 


0,0.3 


0,0 




Overstory % Cover, x + SD 


18.8 + 33.8 


27.4 + 33.5 


51.9 + 42.8 


15.4 + 20.8 




% Low Level Light Available,* +SD b 


67.4 + 34.0 


49.0 + 32.5 


43.8 + 45.4 


58.8 + 33.3 




% High Level Light Available, R +SD C 


73.4 + 30.5 


67.8 + 29 7 


50.8 + 47.8 


77.4 + 28.6 










1 



Table 6-4--continued 



599 















Bidens 


Drosera 


Brasenia 


Lycopodium 


Parameter 


mitis 


intermedia 


schreberi 


spp. 


Sample Size 


59 


35 


13 


22 


Water Depth x + SD (m) 


0.15 + 0.08 


0.17 + 0.09 


0.23+0.06 


0.15 + 0.07 


Minimum Water Depth (m) 


-0.17 


0.00 


0.13 


0.07 


Maximum Water Depth (m) 


0.36 


0.35 


0.30 


0.40 


x %TimeinDCl(SD)" 


8.6 (14.8) 


7.4 (10.0) 


1.2(2.2) 


5.2(4.2) 


x % Time in DC2 (SD) 


7.4(6.3) 


8.1 (6.9) 


2.2 (2.8) 


8.3 (5.5) 


x % Tune in DC3 (SD) 


35.3(17.1) 


31.9(21.2) 


15.9(15.1) 


42.1(13.1) 


x % Time in DC4 (SD) 


41.0(20.1) 


42.2(22.1) 


60.8(11.1) 


37.6(15.8) 


x % Time in DC5 (SD) 


7.0(11.8) 


12.0(15.8) 


19.9(16.5) 


6.8(19.2) 


x % Time in DC6 (SD) 


0.4(2.7) 


0.5(1.5) 


0(0) 


0(0.1) 


x % Time m DC7 (SD) 


0.2(1.5) 


0(0.1) 


0(0) 


0(0) 


Mode of %Time in DC1 











7.3 


Mode of %Time in DC2 











11.1 


Mode of %Time in DC3 


50.6 


57.8 


1.2 


49.6 


Mode of %Time in DC4 


26.1 


16.5 


35.7 


30.3 


Mode of %Time in DC5 


1.0 


0.4 


12.4 


1.7 


Mode of %Time in DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC1 


0, 78.6 


0, 50.9 


0,7.3 


0, 17.9 


Minimum, Maximum % Time in DC2 


0, 35.7 


0,22.8 


0,8.4 


0, 24.0 


Minimum, Maximum % Time in DC3 


1.2,58.2 


2.2, 59.6 


1.2,50.2 


0, 53.4 


Minimum, Maximum % Time in DC4 


1.6,76.4 


9.2, 79.9 


35.7, 76.4 


7.2, 70.4 


Minimum, Maximum % Time in DC5 


0, 74.0 


0,53.3 


1.8,46.2 


0.5, 92.3 


Minimum, Maximum % Time in DC6 


0, 20.7 


0,7.9 


0,0 


0,0.4 


Minimum, Maximum % Time in DC7 


0,11.4 


0,0.4 


0,0 


0,0 


Overstory % Cover, x + SD 


11.2 + 24.1 


18.7 + 27.5 


0.12 + 0.4 


3.7 + 7.1 


% Low Level Light Available, x +SD 


60.7 + 30.7 


58.1+33.5 


76.5 + 18.8 


65.2 + 29.7 


% High Level Light Available,? +SD° 


81.8 + 25.9 


81.6 + 23.6 


89.8+11.0 


91.7+10.3 



Table 6-4--continued. 



600 



















Ludmgia 


Itea 


Smilax 


Smilax 




Parameter 


alata 


virginica 


walteri 


laurifolia 


Sample Size 


6 


95 


38 


49 




Water Depth x + SD (m) 


0.29 + 0.09 


0.18 + 0.27 


0.06 + 0.19 


0.09 + 0.14 




Minimum Water Depth (m) 


0.21 


-0.45 


-0.40 


-0.29 




Maximum Water Depth (m) 


0.42 


1.01 


0.32 


0.63 




x %TimeinDCl(SD)' 


29.0 (6.2) 


26.6 (25.2) 


34.7 (29.2) 


25.0(24.7) 




x % Time in DC2 (SD) 


3.6 (0.4) 


6.5(4.3) 


7.7(4.3) 


10.2 (6.3) 




x % Time in DC3 (SD) 


7.4 (0.3) 


16.5(10.8) 


17.9(10.4) 


31.7(18.1) 




x % Time in DC4 (SD) 


11.9(0.7) 


21.4(11.8) 


22.4(11.8) 


25.0(16.2) 




x % Time in DC5 (SD) 


20.1 (1.5) 


18.5(14.2) 


15.4(14.3) 


5.4 (6.9) 




x % Time in DC6 (SD) 


18.2(2.8) 


6.5(10.0) 


1.6(2.3) 


1.4(4.7) 




x % Time in DC7 (SD) 


9.9 (3.0) 


4.0(10.4) 


0.3 (0.9) 


1.3(5.2) 




Mode of %Time in DC 1 


19.4 


25.0 


26.4 


6.7 




Mode of %Time in DC2 


3.7 


2.6 


13.0 


10.1 




Mode of %Time in DC3 


7.3 


27.5 


25.9 


49.1 




Mode of %Time m DC4 


12.3 


25.9 


25.2 


32.1 




Mode of %Time in DC5 


18.9 


0.9 


9.1 


1.9 




Mode of %Time in DC6 


15.3 


0.9 


0.5 







Mode of %Time in DC7 


8.1 













Minimum, Maximum % Time in DC 1 


19.4,35.1 


0.5, 96.6 


0, 95.0 


0, 96.8 




Minimum, Maximum % Tune in DC2 


2.9, 4.0 


0.2, 16.3 


0, 14.3 


0.1,35.7 




Minimum, Maximum % Time in DC 3 


7.1,8.0 


1 4,55.1 


0.4,49.1 


0.8, 59.6 




Minimum, Maximum % Time in DC4 


10 7,12.7 


0.5, 43.6 


1.0,38.6 


0.1,80.5 




Minimum, Maximum % Time in DC5 


18.9,22.6 


0.2, 57.7 


0.5, 66.3 


0, 29.9 




Minimum, Maximum % Time in DC6 


15.3,22.4 


0, 33.6 


0, 10.1 


0, 22.6 




Minimum, Maximum % Tune in DC7 


7.4,13.8 


0, 48.6 


0,4.2 


0, 32.3 




Overstory % Cover, x + SD 


16.5 + 33.9 


47.3 + 38.2 


60.2 + 37.5 


47.9 + 40.2 




% Low Level Light Available,! +SD 


57.7 + 23.8 


33.9 + 28.9 


27.8 + 28.9 


33.7 + 34.7 




% High Level Light Available,! +SD C 


76.0 + 9.8 


48.3 + 33.1 


37.3 + 36.0 


53.6 + 38.0 















Table 6-4~continued 



601 















Cephalanthus 


Clethra 


Cyrilla 


Pierus 


Parameter 


occidentalis 


alnifolia 


racemiflora 


phillyreifolia 


Sample Size 


15 


12 


119 


34 


Water Depth x + SD (m) 


0.20 + 0.21 


0.21+0.17 


0.11+0.15 


0.11+0.15 


Minimum Water Depth (m) 


-0.38 


0.01 


-0.29 


-0.28 


Maximum Water Depth (m) 


0.50 


0.65 


0.65 


0.63 


x %TimeinDCl (SD)" 


33.8(15.0) 


14.4(14.1) 


25.2 (22.3) 


26.5 (23.2) 


x % Tune in DC2 (SD) 


4.7(3.3) 


7.6 (5.7) 


8.4 (5.4) 


9.9 (6.2) 


x % Time in DC3 (SD) 


9.4 (5.5) 


23.4(14.3) 


22.9(13.8) 


24.6(14.4) 


x % Time in DC4 (SD) 


12.8 (5.8) 


31.9(16.5) 


26.6(15.2) 


25.6(14.8) 


x % Time in DCS (SD) 


17.4(6.8) 


17.4(16.0) 


14.7(18.3) 


11.2(13.0) 


x % Time m DC6 (SD) 


13.6(7.7) 


2.6(4.9) 


1.5(3.2) 


1.3(3.1) 


x % Time in DC7 (SD) 


8.3 (6.0) 


2.8(9.7) 


0.6 (3.4) 


1.0(5.5) 


Mode of %Time m DC 1 


33.1 


0.7 





26.4 


Mode of %Time in DC2 


2.8 


2.0 





13.0 


Mode of %Time in DC3 


7.5 


4.6 


25.9 


22.6 


Mode of %Time m DC4 


12.3 


7.1 


25.2 


25.2 


Mode of %Time in DCS 


19.4 


0.8 


0.8 


9.1 


Mode of %Time in DC6 


0.4 











Mode of %Time in DC7 














Minimum, Maximum % Time in DO 


15.2,75.3 


0.7, 50.0 


0,91.1 


0, 94.8 


Minimum, Maximum % Time in DC2 


2.1, 14.1 


2.0, 19.7 


0,27.5 


0.1,35.7 


Minimum, Maximum % Time in DC3 


5.0, 24.2 


4.6,49.1 


0, 59 6 


2.0,53.6 


Minimum, Maximum % Time in DC4 


5.2, 27.0 


7 1,67.9 


1.4,79.9 


0.7, 80.5 


Minimum, Maximum % Time in DC5 


5.9, 29.8 


0.8, 52.8 


0.1,91.8 


0,51.9 


Minimum, Maximum % Time in DC6 


0.4, 23.6 


0, 17.2 


0, 17.24 


0, 17.0 


Minimum, Maximum % Time in DC 7 


0, 18.2 


0,33.7 


0, 33.7 


0, 32.3 


Overstory % Cover, x + SD 


32.9 + 43.8 


51.2 + 41.8 


42.11+38.9 


62.1+38.0 


% Low Level Light Available,? +SD b 


32.0 + 34.4 


22.3 + 29.0 


39.1+36.8 


18.6 + 24.3 


% High Level Light Available,? +SD C 


483 + 349 


49.5 + 33.5 


55.9 + 37.4 


37.9 + 35.9 



Table 6-4--continued 



602 















Lyonia 


Leucothoe 


Gordonia 


Ilex 


Parameter 


lucida 


racemosa 


lasianthus 


cassine 


Sample Size 


93 


68 


33 


57 


Water Depth x + SD (m) 


0.07 + 0.11 


0.07 + 0.11 


0.08 + 0.10 


0.12 + 0.16 


Minimum Water Depth (m) 


-0.40 


-0.36 


-0.10 


-0.35 


Maximum Water Depth (m) 


0.30 


0.38 


0.32 


0.58 


x % Time in DC 1 (SD)' 


28.1(24.3) 


28.1 (23.6) 


24.9(21.8) 


24.4 (20.8) 


x % Time in DC2 (SD) 


10.1 (5.8) 


9.5(3.9) 


9.8 (5.0) 


8.1 (4.9) 


x % Time in DC3 (SD) 


27.3 (14.7) 


28.9(15.0) 


28.0(13.5) 


20.7(11.8) 


x % Time in DC4 (SD) 


26.2(15.4) 


24.8(11.7) 


27.3(13.1) 


25.8(13.0) 


x % Time in DC5 (SD) 


7.8(9.3) 


7.0 (6.6) 


7.9(11.2) 


15.6(14.4) 


x % Time in DC6 (SD) 


0.4 (0.9) 


1.1(3.4) 


0.2 (0.5) 


2.5 (5.0) 


x % Time in DC7 (SD) 


0(0) 


0.5(2.3) 


0(0) 


1.2(4.6) 


Mode of %Time in DC 1 


6.7 


6.7 


5.2 


7.2 


Mode of %Time in DC2 


10.1 


10.1 


10.5 


9.4 


Mode of %Time in DC3 


17.4 


49.1 


19.3 


14.8 


Mode of %Time in DC4 


32.1 


32.1 


7.9 


30.7 ' 


Mode of %Time in DC5 


0.5 


1.9 


0.9 


0.8 


Mode of %Time in DC6 














Mode of %Time in DC7 














Minimum, Maximum % Time in DC1 


0,96.8 


1.5,93.5 


0.2,63.7 


0,90.1 


Minimum, Maximum % Time in DC2 


0.1,35.7 


1.9,27.5 


0.4,27.3 


0, 24.0 


Minimum, Maximum % Time in 1X3 


0.8, 59.6 


2.4,53.6 


4.1,49.6 


0.2, 49.6 


Minimum, Maximum % Time in DC4 


0.1,80.5 


1.4,60.7 


7.9, 60.9 


0.8,60.9 


Minimum, Maximum % Time in DC5 


0, 45.2 


0.4, 23.9 


0.6, 60.8 


0.3, 76.5 


Minimum, Maximum % Time in DC6 


0,5.3 


0, 20.5 


0,2.4 


0,23.5 


Minimum, Maximum % Time in DC 7 


0,0.3 


0, 12.9 


0,0.2 


0, 30.6 


Overstory % Cover, x + SD 


61.7 + 36.9 


54.0 + 39.2 


d 


d 


% Low Level Light Available, x +SD 


22.0 + 28.8 


29.2 + 31.9 






% High Level Light Available, x +SD C 


39.4 + 36.6 


43.8 + 37.9 







Table 6-4— continued 



603 















Magnolia 


Persea 


Pinus 


Nyssa 


Parameter 


virginiana 


palustris 


spp. 


sylvatica 
v. biflora 


Sample Size 


20 


7 


12 


34 


Water Depth x + SD (m) 


0.11+0.13 


0.11+0.09 


0.15 + 0.11 


0.22 + 0.21 


Minimum Water Depth (m) 


-0.07 


-0.06 


-0.03 


-0.32 


Maximum Water Depth (m) 


0.44 


0.21 


0.37 


0.63 


x % Time in DC 1 (SD)' 


26.9(19.6) 


25.6(19.3) 


13.5(16.0) 


26.5(18.6) 


x % Time in DC2 (SD) 


9.7(5.2) 


8.0(2.9) 


11.0(8.1) 


4.3(3.2) 


x % Time in DC3 (SD) 


23.0(12.1) 


20.5 (3.6) 


27.0(15.5) 


10.4 (6 4) 


x % Time m DC4 (SD) 


24.4 (9.8) 


30.1 (15.3) 


30.1 (13.1) 


18.2(11.8) 


x % Time in DC5 (SD) 


12.8(13.2) 


14.8(7.3) 


18.2(29.9) 


22.3(11.8) 


x % Time in DC6 (SD) 


2.3 (5.0) 


1.0(0.6) 


0(0) 


10.3 (76) 


x % Time in DC7 (SD) 


0.9(3.5) 


0(0) 


0(0) 


7.4 (8.9) 


Mode of %TimeinDCl 


3.8 


1.6 





23.6 


Mode of %Time m DC2 


2.2 


3.1 





3.9 


Mode of %Time in DC3 


7.0 


14.8 


0.1 


7.4 


Mode of %Time in DC4 


9.8 


12.9 


13.9 


10.9 


Mode of %Time in DCS 


0.6 


4.4 


5.2 


19.2 


Mode of %Time m DC6 











1.8 


Mode of %Time in DC7 


o 











Minimum. Maximum % Time in DC 1 


3.8,61.2 


1.6,59.0 


0,51.9 


1.9,71.8 


Minimum, Maximum % Time in DC2 


2.2,11.9 


3.1,11.4 


0, 27.3 


1.5,18.1 


Minimum, Maximum % Time in DC3 


7.0, 46.9 


14.8, 23.9 


0.1,45.6 


3.9,25.1 


Minimum, Maximum % Time in DC4 


9.8, 40.1 


12.9,60.9 


13.9,51.2 


5.9,38.3 


Minimum, Maximum % Time in DC5 


0.6, 46.2 


4.4, 26.6 


0.8, 85.9 


7.3, 59.3 


Minimum, Maximum % Time in DC6 


0, 22.0 


0, 17 


0,0.1 


4,23.5 


Minimum, Maximum % Time in DC7 


0, 15.5 


0,0.1 


0,0 


0, 32.6 


Overstory % Cover, x + SD 


d 


d 


d 


d 



Table 6-4--continued. 



604 















Acer 


Taxodium 


Nyssa 


Ilex 


Parameter 


rubrum 


ascendens 


ogeechee 


myrtifolia 


Sample Size 


8 


93 


6 


3 


Water Depth x + SD (m) 


0.08 + 0.19 


0.17 + 0.18 


0.10 + 0.21 


0.09 + 0.33 


Minimum Water Depth (m) 


-0.22 


-0.35 


-0.22 


-0.22 


Maximum Water Depth (m) 


0.44 


0.63 


0.32 


0.44 


x %TimeinDCl (SD)* 


40.0(17.9) 


22.2 (20.5) 


42.9(14.3) 


43.8(21.2) 


x % Time in DC2 (SD) 


4.4(3.3) 


5.9(4.1) 


3.0 (0.8) 


2.6 (0.5) 


x % Time in DC3 (SD) 


9.8 (7.6) 


15.5(10.1) 


6.0 (2.0) 


5.1(1.6) 


x % Time in DC4 (SD) 


13.6(10.7) 


25.5(13.1) 


8.7 (3.0) 


7.6(2.7) 


x % Time in DC5 (SD) 


15.4(6.4) 


23.0(17.8) 


15.5(5.0) 


14.5(6.1) 


x % Time m DC6 (SD) 


9.7 (6.9) 


4.3 (5.9) 


13.9(4.2) 


14.2(7.1) 


x % Tune in DC7 (SD) 


7.2 (5.9) 


2.6 (6.4) 


10.1 (2.3) 


12.2(3.8) 


Mode of %Time in DC1 


9.2 





26.6 


20.2 


Mode of %Time in DC2 


2.1 


2.2 


2.1 


2.2 


Mode of %Time in DC3 


3.9 


7.4 


3.9 


3.9 


Mode of %Time in DC4 


5.9 


10.9 


5.9 


5.9 


Mode of %Time in DC5 


8.0 


19.2 


10.4 


10.4 


Mode of %Time in DC6 


0.4 





8.3 


8.3 


Mode of %Time in DC7 








6.7 


8.1 


Minimum, Maximum % Time in DC1 


9.2,61.3 


0,90.1 


26.6,61.3 


20.2,61.3 


Minimum, Maximum % Time in DC2 


2.1,12.1 


0,18.1 


2.1,3.9 


2.2,3.1 


Minimum, Maximum % Time in DC3 


3.9, 23.3 


0.1,49.6 


3.9, 8.0 


3.9, 7.0 


Minimum, Maximum % Time in DC4 


5.9, 37.0 


0.8, 60.9 


5.9, 12.2 


5.9, 10.7 


Minimum, Maximum % Time in DC5 


8.0, 26.8 


0.3,85.9 


10.4,21.1 


10.4,21.5 


Minimum, Maximum % Time in DC6 


0.4, 22.0 


0, 22.0 


8.3, 19.0 


8.3, 22.0 


Minimum, Maximum % Time in DC7 


0,15.5 


0, 32.6 


6.7,13.1 


8.1,15.5 


Overstory % Cover, x + SD 


d 


d 


d 


d 



Table 6-4~continued. 







Fraxinus 


Parameter 


caroliniana 


Sample Size 


3 


Water Depth x + SD (m) 


0.26 + 0.28 


Minimum Water Depth (m) 


-0.06 


Maximum Water Depth (m) 


0.44 


x % Time in DC 1 (SD)* 


33.7(21.9) 


x % Time in DC2 (SD) 


5.0(3.1) 


x % Time in DC3 (SD) 


9.7(4.4) 


x % Time in DC4 (SD) 


11.5(1.2) 


x % Time in DC5 (SD) 


15.6(9.7) 


x % Time in DC6 (SD) 


14.5(12.3) 


x % Time in DC7 (SD) 


9.9(8.6) 


Mode of %TimeinDCl 


20.2 


Mode of %Time m DC2 


3.1 


Mode of %Time in DC3 


7.0 


Mode of %Time in DC4 


10.7 


Mode of %Time in DCS 


4.4 


Mode of %Time in DC6 


0.3 


Mode of %Time in DC7 





Minimum, Maximum % Time in DC 1 


20.2, 59.0 


Minimum, Maximum % Time in DC2 


3.1,8.6 


Minimum, Maximum % Time in DC3 


7.0, 14.8 


Minimum, Maximum % Time in DC4 


10.7, 12.9 


Minimum, Maximum % Time in DC5 


4 4,21.5 


Minimum, Maximum % Time in DC6 


0.3, 22.0 


Minimum, Maximum % Time in DC7 


0, 15.5 


Overstory % Cover, x + SD 


d 



605 



Table 6-4--continued. 



606 



a Depth Classes (DC) are: 



(no inundation) DC 1 
(shallow water) DC 2 
(shallow water) DC 3 
(shallow water) DC 4 
(deep water) DC 5 
(deep water) DC 6 
(deep water) DC 7 



water depth < 0.0 m 
0.00 m < water depth < 0.05 m 
0.05 m < water depth < 0. 15 m 
0. 15 m< water depth < 0.30 m 
0.30 m < water depth < 0.60 m 
0.60 m < water depth < 1.00 m 
water depth > 1 .00 m 



b Light availability measured at 0.3 m above the peat surface with a Licor quantum 
sensor. See chapter text for details. 

c Light availability measured at 1 .0 m above the peat surface with a Licor quantum 
sensor. See chapter text for details. 

d Overstory % cover and availability of light at 0.3 m and 1 .0 m above the peat surface 
were not estimated for tree belt transects. 






607 

Table 6-5. Mest (mean water depth) and Wilcoxon rank-sum (percent of interval in each 
depth class) comparisons of hydrologic environments during 1962-1995 where species 
were present and absent in vegetation sample plots during 1993-1994. 



Parameter 


Carex 
waheriana 


Nymphaea 
odorata 


Xyris spp. 


Vtricuhuw 
ipp. 


Sample Size, Species Present 
Sample Size. Species Absent 


437 
505 


361 
583 


248 
694 


244 
698 


I Water Depth (SD) (mX Present 
1 Water Depth (SD)(mX Absent 
P>t 


0.17(0.14) 

0.22 (0.25) 

0.0003 


0.24(0.13) 

0.17(0.24) 

0.0001 


0.17(0.12) 

0.21 (0.23) 

0.0002 


0.29(0.13) 

0.17(0.22) 

0.0001 


i %Time in DC1 (SD), Present 
1 %Time in DO (SD). Absent 
P>Z 


17.7 (20.8) 

22.1 (23.0) 

0.0216 


11.4(16.0) 

25.4(23.6) 

0.0001 


13.0(18.9) 

22.6 (22.7) 

0.0001 


8.9(12.2) 

24.0 (23.4) 

0.0001 


7 % Time in DC2(SDX Present 
H % Time in DC2 (SD). Absent 
P>Z 


6.1(4.3) 
5.3(5.1) 

0.0001 


4.1(3.7) 
6.6(5.1) 
0.0001 


6.6 (5.4) 
5.3(4.5) 

o.ooio 


3.4(2.9) 
6.5 (5.0) 
0.0001 


7 %Time in DC3 (SD). Present 
■ % Time in DC3 (SD). Absent 
P>Z 


18.7(12.5) 

14.7(13.8) 

0.0001 


13.7(12.2) 

18.3(13.8) 

0.0001 


25.1 (16.9) 

13.5(10.3) 

0.0001 


11.2(9.5) 

18.4(14.0) 

0.0001 


7 % Time in DC4 (SD). Present 
X % Time in DC4 (SD). Absent 
P>Z 


30.5(14.2) 

22.8(16.7) 

0.0001 


30.2(15.7) 

24.1(15.8) 

0.0001 


36.0(18.6) 

23.0(13.5) 

0.0001 


29.1(14.4) 

25.5(16.5) 

0.0002 


I % Time in DC5 (SD), Present 
x % Time in DC5 (SD), Absent 
P>Z 


24.8 (20.2) 

25.0 (23.3) 

0.4227 


37.1 (24.2) 

17.3 (16.2) 

0.0001 


18.3(20.3) 

27.2 (22.0) 

0.0001 


41.8(21.6) 

19.0(18.7) 

0.0001 


1 % Time in DC6(SDX Present 
■ % Time in DC6 (SD), Absent 
P>Zt 


2.1(3.9) 
5.9(8.5) 
0.0002 


3.3 (6.3) 
4.6 (7.4) 
0.5824 


1.0(1.8) 
5.3 (7.8) 
0.0001 


4.8 (7.7) 
3.9(6.8) 
0.0001 


t % Time in DC7 (SDX Present 
> %Time in DC7(SDX Absent 
P>Z 


0.1 (0.8) 
4.2(11.6) 

0.0001 


0.2(1.1) 

3.6(10.9) 

0.0002 


0.0(0.1) 

3.1(10.1) 

0.0001 


0.9(5.0) 
2.8(9.7) 
0.8521 


! Overstory % Cover (SD), Present 
■ Overstorv N Cover (SD). Absent 
P>Z 


29.3 (36.8) 

27.4(37.9) 

0.0266 


13.2(27.1) 

37.6(39.8) 

0.0001 


25.5(35.1) 

29.3(38.1) 

0.5831 


20.8(34.1) 

30.9(38.1) 

0.0001 


b 

s % Low Level Light Available (SDX Present 
2 % Low Level Light Available (SDX Absent 
P>Z 


47.6(35.0) 

50.5 (39.2) 

0.1948 


68.6(34.3) 

37.1 (33.9) 

0.0001 


56.5 (35.5) 

46.5 (37.6) 

0.0001 


60.3 (38.8) 

45.3 (36.0) 

0.0001 


s % High Level Light Available (SD). Present 
» % High Level Light Available (SD). Absent 
P>Z 


68.5 (33.3) 

66.3 (35.0) 

0.4431 


82.7 (25.6) 

57.8 (35.4) 
0.0001 


71.0(33.0) 

66.0 (34.6) 

0.0070 


74.6 (30.7) 

64.8 (35.0) 

0.0001 









Table 6-5--continued. 



608 





Eriocaulon 


Khynchosporu 


EUocharis 


Panieum 


Parameter 


lineare 


inundata 


baUhvinii/ 
vtvipara 


hemitomon 


Sample Size, Species Present 
Sample Size, Species Absent 


44 
898 


35 
907 


157 
785 


228 

715 


x Water Depth (SD) (m). Present 
x Water Depth (SD) (m). Absent 
P>t 


0.16(0.08) 

0.20 (0.22) 

0.0046 


0.17(0.13) 

0.20(0.21) 

0.2700 


0.16(0.20) 

0.21(0.21) 

0.0078 


0.19(0.15) 

0.20 (0.23) 

0.6135 


St % Time in DC1 (SD), Present 
! °/o Time in DC 1 (SD). Absent 
P>Z 


5.3 (9.0) 

20.8 (22.3) 

0.0001 


13.6(21.8) 

20.3(22.1) 

0.0029 


30.9(21.5) 

17.9(21.6) 

0.0001 


15.8(20.0) 

21.4(22.6) 

0.0001 


1 % Time in DC2 (SD), Present 
< % Time in DC2(SD), Absent 
P>Z 


6.9(4.7) 
5.6 (4.8) 
0.0385 


5.8 (7.2) 
5.7(4.6) 
0.3704 


6.3 (5.0) 

5.5(4.7) 
0.0225 


5.4(4.8) 
5.7(4.7) 
0.2035 


It % Time in DC3 (SD), Present 
I % Time in DC3 (SD). Absent 
P>Z 


39.3 (19.7) 

15.4(11.9) 

0.0001 


19.8(17.6) 
16.4(13.2) 

0.5770 


16.5(13.3) 

16.6(13.4) 

0.7395 


18.1 (14.9) 

16.1(12.8) 

0.2261 


1 % Time in DC4(SD). Present 
x % Time in DC4 (SD), Absent 
P>Z 


38.5 (19.2) 

25.8(15.7) 

0.0001 


36.4(21.9) 

26.0 (15.7) 

00065 


19.8(15.6) 

27.7(15.8) 

0.0001 


29.9(16.4) 

25.3(15.8) 

0.0006 


x % Time in DC5(SD). Present 
x % Time in DC5(SD). Absent 
P>Z 


9.9 (21.4) 

25.6(21.7) 

<0.0001 


24.1 (25.1) 

24 9(21.8) 

0.3981 


13.8(11.5) 

27.1 (22.8) 
0.0001 


27.5 (24.0) 

24.1(21.1) 

0.1138 


x % Time in DCo(SD). Present 
3 % Time in DC6(SD). Absent 
P>Zt 


0.0(0.1) 
4.3(7.1) 
0.0001 


0.2 (0.6) 
4.3(7.1) 
0.0001 


7.1(8.6) 
3.5(6.5) 

0.0528 


2.8 (5.3) 
4.5 (7.5) 
0.0014 


x % Time in DC7 (SD). Present 
* % Time in DC7(SD). Absent 
P>Z 


0.0 (0.0) 
2.4(9.0) 
0.0001 


0.0 (0.0) 
2.4(8.9) 
0.0001 


5.5 (8.3) 
1.7(8.7) 
0.0001 


0.5 (2.3) 
2.9 (9.9) 

00706 


x Overstory% Cover (SD). Present 
: Overstory % Cover (SD). Absent 
P>Z 


10.9(16.9) 

29.2 (37.9) 

0.0154 


10.9(17.5) 

29.0(37.8) 

0.1278 


21.4 (32.7) 

29.7(38.1) 

0.1166 


16.7(28.5) 

32.0(39.1) 

0.0001 


2 % Low Level Light Available (SD). Present 
7 % Low Level Light Available (SD), Absent 
P>Z 


66.0(33.3) 

48.3 (37.3) 

0.0007 


71.5(30.1) 

48.3 (37.3) 

0.0006 


49.4 (35.0) 

49.1(35.0) 

0.6768 


61.3(34.9) 

45.3 (37.2) 

0.0001 


x % High Level Light Available (SD), Present 
x % High Level Light Available (SD). Absent 
P>Z 


88.0(18.1) 

66.3 (34.5) 

0.0001 


87.4(19.7) 

66.6 (34.4) 

0001 


70.8(29.4) 

66.6(35.1) 

0.4921 




79.7(25.9) 

63.4(35.6) 

0.0001 



Table 6-5-continued. 








609 




Parameter 


Lacnantkes 


Nuphar 


fVoodwardia 


Peltundra 








caroliniana 


luteum 


virginica 


virginica 




Sample Size, Species Present 


266 


68 


197 


356 




Sample Size. Species Absent 


676 


874 


745 


586 






J Water Depth (SD) (m). Present 


0.10(0.16) 


0.26(0.21) 


0.12(0.19) 


0.15(0.14) 






■ Water Depth (SD)(m). Absent 


0.24 (0.22) 


0.19(0.21) 


0.22(0.21) 


0.23 (0.24) 






P>t 


0.0001 


0.0123 


0.0001 


0.0001 






x % Time in DC1 (SD), Present " 


28.5 (25.4) 


23.0 (19.8) 


25.0 (24.2) 


19.8(22.1) 






x % Time in DC 1 (SD), Absent 


16.7 (19.7) 


19.8 (22.3) 


18.8(21.4) 


20.2 (22.2) 






P>Z 


0.0001 


0.0914 


0.0008 


0.8028 






a % Time in DC2(SD), Present 


7.5 (4.7) 


3.5 (2.9) 


8.5 (5.5) 


6.9 (4.7) 






x % Time in DC2 (SD). Absent 


4.9 (4.6) 


5.8 (4.8) 


4.9(4.2) 


4.9 (4.6) 






P>Z 


0.0001 


0.0001 


0.0001 


0.0001 






■ % Time in DC3 (SD). Present 


21.8(15.0) 


11.2(11.3) 


24.9 (15.8) 


21.7(14.4) 






x % Time in DC3 (SD). Absent 


14.5(12.1) 


17.0(13.4) 


14.3(11.7) 


13.4(11.6) 






P>Z 


0.0001 


0.0001 


0.0001 


0.0001 






H % Time in DC4 (SD), Present 


25.0(16.2) 


22.5 (22.8) 


26.0(15.6) 


30.2(15.0) 






s % Time in DC4 (SD), Absent 


27.0(15.9) 


26.7(15.4) 


26.5 (16.2) 


24.1(16.2) 






P>Z 


0.0295 


0.0001 


0.5396 


0.0001 






« % Time in DCS (SD), Present 


13.4(15.4) 


20.6(18.8) 


13.3 (17.9) 


19.8(19.1) 






* % Time in DC5 (SD), Absent 


29.4(22.4) 


25.2(22.1) 


28.0(21.8) 


28.0 (22.9) 






P>Z 


0.0001 


0.2710 


0.0001 


0.0001 






x % Time in DC6 (SD), Present 


2.4(5.2) 


11.0(9.0) 


1.7(4.7) 


1.5(3.3) 






X % Time in DC6(SDX Absent 


4.8(7.5) 


3.6 (6.6) 


4.8(7.4) 


5.7(8.1) 






P>Zt 


0.0001 


0.0001 


0.0001 


0.0001 






x % Time in DC7(SD), Present 


1.3 (4.5) 


8.2 (9.5) 


0.6(6.5) 


0.0(0.1) 






5 % Time in DC7(SDX Absent 


2.7(9.3) 


1.9(8.5) 


2.8 (9.2) 


3.7(10.9) 






P>Z 


0.0002 


0.0001 


0.0001 


0.0001 






; Overstory% Cover (SD). Present 


28.6 (35.6) 


9.6 (26.5) 


39.6(38.1) 


32.8 (36.6) 






H Overstory % Cover (SD), Absent 


28.2(38.1) 


29.7(37.7) 


25.3 (36.6) 


25.6 (37.6) 






P>Z 


0.0300 


0.0001 


0.0001 


0.0001 






x % Low Level Light Available (SD), Present 


45.7(35.3) 


48.9(35.8) 


33.2(33.6) 


44.8 (35.5) 






x % Low Level Light Available (SD), Absent 


50.5 (35.3) 


49.2 (37.4) 


53.4(37.1) 


51.8(38.1) 






P>Z 


0.1714 


0.3987 


0.0001 


0.0168 






x % High Level Light Available (SD), Present ' 


69.0 (33.3) 


77.7(21.1) 


59.4(37.0) 


64.5 (35.0) 






x % High Level Light Available (SD). Absent 


66.6 (34.6) 


66.5 (34.9) 


69.4(33.2) 


69.0(33.7) 






P>Z 


0.2495 


0.6409 


0.0087 


0.0354 









Table 6-5--continued. 



610 



Parameter 



Hidens 
mitis 



Drosera 
intermedia 



Brosenia 
schreberi 



l.ycopodium 
spp. 



Sample Size. Species Present 
Sample Size. Species Absent 



I Water Depth (SD)(m). Present 
I Water Depth (SD) (m). Absent 
P>t 



1 % Time in DC 1 (SD). Present 
; % Time in DC 1 (SD). Absent 
P>Z 



t % Time in DC2 (SD). Present 
• % Time in DC2(SD), Absent 
P>Z 



* % Time in DC3 (SD), Present 

• "iTime in DC3 (SD). Absent 
P>Z 



J % Time in DC4(SD). Present 
I % Time in DC4(SD), Absent 
P>Z 



« % Time in DCS (SD). Present 
■ % Time in DCS (SD). Absent 
P>Z 



s % Time in DC6 (SD). Present 
« % Time in DC6 (SD). Absent 
P>Zl 



B % Time in DC7(SDX Present 
; % Time in DC7 (SD). Absent 
P>Z 



* Oventory% Cover (SD). Present 
■ Overstory % Cover (SD). Absent 
P>Z 



* % Low Level Light Available (SD). Present 
I % Low Level Light Available (SD). .Absent 
P>Z 



; % High Level Light Available (SD). Present 
; % High Level Light Available (SD), Absent 
P>Z 



S9 

885 



0.15(0.08) 

0.20 (0.22) 

0.0001 



8.6(14.8) 

20.8 (22.3) 

0.0001 



7.4 (7.3) 

5.5 (4.6) 
0.0149 



35.3(17.1) 

15.3(12.1) 

0.0001 



41.0(20.1) 

25.4(15.3) 

0.0001 



7.0(11.8) 

26.1(21.9) 

0.0001 



0.4(2.7) 
4.4 (7.2) 
0.0001 



0.2(1.5) 
2.5 (9.0) 
0.0001 



11.2(24.1) 

29.4(24.1) 

0.0003 



60.7(30.7) 

48.4 (37.6) 

0.0166 



81.8(25.9) 

66.4(34.5) 

0.0001 



35 

907 



0.17(0.09) 

0.20(0.21) 

0.0469 



7.4(10.0) 

20.5 (22.3) 

0.0001 



8.1(6.9) 
5.6(4.6) 
0.0972 



31.9(21.2) 

16.0(12.6) 

0.0001 



40.2(22.1) 

25.9(15.5) 

00004 



12.0(15.8) 

25.4(21.9) 

0.0001 



0.5(1.5) 
4.3(7.1) 
0.0001 



0.0(0.1) 
2.4(8.9) 
0.0009 



18.7(27.5) 

28.7(37.7) 

0.2724 



58.1 (33.5) 

48.8 (37.4) 

0.0979 



81.6(23.6) 

66.8 (34.5) 

0.0150 



13 
929 



0.23 (0.06) 

0.20(0.21) 

0.1220 



1.2(2.2) 

20.3 (22.2) 

0.0001 



2.2 (2.8) 
5.7(4.8) 
00006 



15.9(15.1) 

16.6(13.4) 

0.6802 



60.8(11.1) 

25.9(15.6) 

0.0001 



19.9(16.5) 

25.0(22.0) 

0.6017 



0.0(0.0) 
4.1(7.1) 
0.0001 



0.0 (0.0) 
2.3 (8.8) 
0.0177 



0.1 (0.4) 

28.7 (37.5) 

0.0002 



76.5(18.8) 

48.8 (37.4) 

0.0185 



89.8(110) 

67 (34.3) 

0.0134 



22 
922 



0.15(0.06) 

0.20(0.21) 

0.0019 



5.2 (4.2) 

20 4(22.3) 

0.0006 



8.3 (5.5) 
5.6(4.7) 
0.0079 



42.1 (13.1) 

15.9(12.8) 

0.0001 



37.6(15.8) 

26.1 (16.0) 

0.0005 



6.8(19.2) 

25.3(21.8) 

0.0001 



0.0(0.1) 
4.2(7.1) 
0.0001 



0.0 (0.0) 
2.4 (8.9) 
0.0019 



3.7(7.1) 

28.9 (37.6) 

0.0283 



65.2 (29.7) 

48.8 (37.4) 

0.0271 



917(10.3) 

66.7 (34.4) 

0.0001 









Table 6-5-continued. 



611 



Parameter 



Sample Size, Species Present 
Sample Size, Species Absent 



* Water Depth (SD) (m), Present 
S Water Depth (SD) (m). Absent 
P>t 



B % Time in DC1 (SD), Present 
B % Time in DC1 (SD), Absent 
P>Z 



B % Time in DC2 (SD). Present 
2 % Time in DC2 (SD). Absent 
P>Z 



B % Time in DC3 (SD). Present 
S % Time in DC3 (SD). Absent 
P>Z 



5 % Time in DC4 (SD), Present 
B % Time in DC4 (SD), Absent 
P>Z 



J % Time in DC5 (SD), Present 
B % Time in DC5(SDX Absent 
P>Z 



J % Time in DC6(SD), Present 
B % Time in DC6(SD), Absent 
P>Zt 



B % Time in DC7(SD), Present 
J % Time in DC7 (SD). Absent 
P>Z 



; Overstory% Cover (SDX Present 
B Overstorv % Cover (SD), Absent 
P>Z 



• % Low Level Light Available (SD), Present 
B % Low Level Light Available (SD), Absent 
P>Z 



B % High Level Light Available (SD). Present 
B % High Level Light Available (SDX Absent 
P>Z 



Sphagnum 
spp. 



310 
634 



0.18(0.26) 

0.21 (0.18) 

0.0441 



20.4(21.9) 

19.9(22.3) 

0.9509 



7.7(5.7) 
4.7(3.8) 
0.0001 



24.4 (16.8) 

12.8(9.2) 

0.0001 



28.6(18.9) 

25.3 (14.4) 

0.2937 



13.4(17.4) 

30.5(21.7) 

0.0001 



2.5 (6.2) 
4.9 (7.3) 
0.0001 



3.1(13.1) 
1.9(5.5) 
0.0001 



24.7 (34.0) 

30.1(38.8) 

0.3263 



49.8 (36.3) 

48.8 (37.8) 

0.4174 



71.8(32.6) 

65.2 (34.8) 

0.0006 



Andropogon 
virginica 



61 

881 



0.17(0.06) 

0.20(0.22) 

0.0008 



5.8 (5.8) 

21.1(22.5) 

0.0001 



6.5(4.1) 
5.6 (4.8) 
0.0337 



34.9(16.4) 

15.3(12.2) 

0.0001 



42.2 (16.6) 

25.3(15.4) 

0.0001 



10.2 (14.0) 

25.9 (22.0) 

0.0001 



0.4(1.0) 
4.4 (7.2) 
0.0001 



0.0 (0.0) 
2.5 (9.0) 
0.0001 



8.2(19.5) 

29.7 (37.9) 

0.0001 



68.1 (28.7) 

47.8 (37.5) 

0.0001 



88.4(18.0) 

65.8 (34.6) 

0.0001 



Dulichium 
arendinaceum 



162 
780 



0.18(0.24) 

0.20 (0.20) 

0.3078 



20.6 (20.3) 

20.0 (22.5) 

0.2533 



6.8(4.1) 
5.4(4.9) 
0.0001 



21.7(13.7) 

15.5(13.1) 

0001 



29.8(17.2) 

25.7(15.7) 

0.0121 



15.7(14.9) 

26.8 (22.6) 

0.0001 



2.7(5.7) 
4.4 (7.2) 
0.0001 



2.8(12.7) 
2.2 (7.7) 
0.0086 



26.8 (35.7) 

28.6 (37.7) 

0.6994 



49.6 (35.2) 

49.1 (37.7) 

0.9247 



72.1 (31.1) 

66.3 (34.8) 

0.1117 



Orontium 
aquaticum 



133 
809 



0.26(0.12) 

0.19(0.22) 

0.0001 



10.2 (12.9) 

21.7(22.9) 

0.0001 



4.2 (4.6) 
5.9 (4.7) 
0.0001 



12.8(12.2) 

17.2(13.5) 

0.0001 



28.0(14.1) 

26.1 (16.3) 

0.1290 



42.3 (25.5) 

22.0(19.8) 

0.0001 



2.6 (5.6) 
4.4 (7.2) 
0.7265 



0.1 (0.2) 
2.7(9.4) 
0.0001 



12.0(25.3) 

31.0(38.4) 

0.0001 



67.2 (37.3) 

46.2 (36.5) 

0.0001 



84.5 (25.5) 

64.5 (34.7) 

0.0001 






Table 6-5--continued. 



612 



Parameter 



Saggetaria 
graminea 



Tridenum 
virginicum 



Sarracenia 
/lava 



Sarracenia 
psittacenia 



Sample Size, Species Present 
Sample Sue, Species Absent 

■ Water Depth (SD) (m). Present 
I Water Depth (SD) (m). Absent 
P>t 



a % Time in DC1 (SD). Present 
i % Time in DC1 (SD). Absent 
P>Z 



■ % Time in DC2(SD), Present 
a % Time in DC2 (SD). Absent 
P>Z 



I % Time in DC3 (SD). Present 
I % Time in DC3(SDX Absent 
P>Z 



■ °/o Time in DC4 (SD). Present 

■ % Time in DC4(SD), Absent 
P>Z 



a % Time in DCS (SD), Present 
2 % Time in DC5 (SD). Absent 
P>Z 



a % Time in DC6 (SD). Present 
a % Time in DC6 (SD). Absent 
P>Zt 



K % Time in DC7(SD), Present 
a % Time in DC7 (SD), Absent 
P>Z 



» Overstory% Cover (SD). Present 
R Overstory % Cover (SD). Absent 
P>Z 



J % Low Level Light Available (SDX Present 
■ % Low Level Light Available (SD). Absent 
P>Z 



a % High Level Light Available (SD). Present 
- % High Level Light Available (SD). Absent 
P>Z 



66 
878 

0.15(0.10) 

0.20 (0.22) 

0.0006 



14.9(18 8) 

20.5 (22.3) 

0.0576 



6.9 (4 6) 
5.6(4.8) 
0.0038 



26.8(16.2) 

15.8(12.8) 

0.0001 



35.3 (16.9) 

25.7(15.8) 

0.0001 



15.4(16.1) 

25.6(22.1) 

0.0002 



0.6(1.0) 
4.4(7.2) 
0.0001 



0.0 (0.0) 
2.5(9.1) 
0.0001 



15.3(26.5) 

29.3 (37.9) 

0.0402 



62.0(30.6) 

48.2 (37.6) 

0.0075 



80.8 (23.7) 

66.3 (34.7) 

0.0016 



42 
902 

0.14(0.08) 

0.20 (0.22) 

0.0001 



7.9(16.4) 

20.6(22.2) 

0.0001 



6.7(4.5) 
5.6 (4.8) 
0.0729 



36.6(17.0) 

15.6(12.4) 

00001 



43.0 (20.2) 

25.6 (15.4) 

0.0001 



5.6(7.7) 

25.8(21.9) 

0.0001 



0.1 (0.6) 
4.3(7.1) 
0.0001 



0.0 (0.0) 
2.4(9.0) 
0.0001 



3.7(10.6) 

29.4(37.8) 

0.0001 



70.6(27.3) 

48.2 (37.4) 

0.0001 



89.9(17.1) 

66.3 (34.5) 

0.0001 



62 
882 

0.13(0.10) 

0.20 (0.22) 

0.0001 



10.1(18.1) 

20.7(22.2) 

0.0001 



8.6 (5.8) 
5.6 (4.6) 
0.0001 



38.8(18.9) 

15.0(11.4) 

0.0001 



33.9(18.1) 

25.9(15.8) 

0.0003 



8.5(19.4) 

26.0(21.6) 

0.0001 



0.0(0.2) 
4.4(7.2) 
0.0001 



0.0(0.0) 
2.4 (9.0) 
0.0001 



12.6(16.7) 

29.4 (38.2) 

0.3523 



54.6(31.4) 

48.8 (37.7) 

0.1237 



80.5 (24.9) 

66.4(34.6) 

0.0013 



15 
927 

0.12(0.05) 

0.20(0.21) 

0.0001 



8.4(13.5) 

20.2 (22.2) 

0.0231 



8.4(3.1) 
5.6(4.8) 
0.0021 



47.8 (9.9) 

16.1 (12.8) 

0.0001 



33.5 (9.7) 

26.3(16.1) 

0.0069 



1.9(1.4) 

25.3(21.9) 

0.0001 



0.0(0.0) 
4.2(7.1) 
0.0001 



0.0(0.0) 
2.4(8.3) 
0.0107 



14.3 (22.6) 

28.6 (37.6) 

0.5914 



48.8(27.5) 

49.2 (37.5) 

0.9331 



83.1 (28.1) 

67.0 (34.3) 

0.0069 



Table 6-5--continued. 



613 



Parameter 



Eleocharis 
robbinsii 



Iris 
virgiitiana 



Decodon 
verticiliatus 



Rhynchospora 

ckalerocephala/ 

wrigktiana 



Sample Size, Species Present 
Sample Size. Species Absent 

t Water Depth (SD) (m). Present 
a Water Depth (SD) (m). Absent 
P>t 



x %Time in DO (SD), Present 
X" % Time in DC1 (SD), Absent 
P>Z 



8 % Time in DC2 (SD), Present 
x •/• Time in DC2 (SD), Absent 
P>Z 



9 % Time in DC3(SDX Present 
• % Time in DC3 (SD). Absent 
P>Z 



x % Time in DC4(SD), Present 
8 % Time in DC4 (SD). Absent 
P>Z 



x % Time in DCS (SD), Present 
I % Time in DC5(SD). Absent 
P>Z 



R % Time in DC6(SD). Present 
■ "/.Time in DC6(SD). Absent 
P>Zt 



■ % Time in DC7 (SD). Present 
x % Time in DC7 (SD). Absent 
P>Z 



x Overstory*. Cover (SD), Present 
x Overstory % Cover (SD), Absent 
P>Z 



x % Low Level light Available (SD), Present 
* % Low Level Light Available (SD). Absent 
P>Z 



a % High Level Light Available (SD), Present 
x % High Level Light Available (SD), Absent 
P>Z 



91 
851 

0.30(0.07) 

0.19(0.22) 

0.000 1 



6.3 (4.5) 

21.5 (22.7) 

0.0001 



3.0(1.6) 
5.9 (4.9) 
0.0001 



10 1(4.4) 

17.2(13.8) 

0.0001 



30.7(7.2) 

25.9 (16.6) 

0.0001 



44.5(9.5) 

22.8(21.8) 

0.0001 



5.3 (6.0) 
4.0(7.1) 
0.0001 



0.2 (0.3) 
2.5 (9.2) 
0.0009 



18.8(33.8) 

29.3 (37.6) 

0.0024 



67.4(34.0) 

47.2 (34.0) 

0.0001 



73.4 (30.5) 

66.7 (34.5) 

0.3230 



43 
899 

0.18(0 08) 

0.20(0.22) 

0.1036 



11.6(10.8) 

20.5 (22.4) 

0.1866 



7.8(5.2) 
5.6 (4.7) 
0.0025 



26.7(14.2) 
16.1(13.1) 

0.0001 



33.7(12.2) 

26.1(16.1) 

0.0001 



18.8(15.6) 

25.2(22.1) 

0.1350 



1.4(1.6) 
4.3 (7.2) 
0.0806 



0.0(0.1) 
2.4 (9.0) 
0.0909 



27.4(33.5) 

28.3 (37.6) 

0.2301 



49.0(32.5) 

49.2 (37.5) 

0.7768 



67.8 (29.7) 

67J (34.4) 

0.3708 



26 
916 

0.19(0.14) 

0.20(0.21) 

0.7533 



10.9(16.5) 

20.3 (22.2) 

0.0046 



6.6 (7.6) 
5.6(4.7) 
0.9438 



25.1(18.7) 

16.3(13.1) 

0.0349 



37.6(23.6) 

26.1(15.7) 

0.0202 



15.3(17.0) 

25.2 (22.0) 

0.0208 



3.9(8.6) 
4.1(7.0) 
0.0030 



0.5(1.3) 
2 4(8.9) 
0.4237 



20.5 (34.0) 

28.5 (37.5) 

0.3314 



54.7(35.0) 

49.0 (37.4) 

0.5651 



72.6 (34.9) 

67.2 (34.2) 

0.2912 



44 
898 

0.10(0.14) 

0.20(0.21) 

0.0001 



20.1 (27.8) 

20.1 (21.8) 

0.1007 



7.7(5.7) 
5.6(4.7) 
0.0149 



30.3(18.7) 

15.9(12.7) 

0.0001 



33.6(23.0) 

26.0(15.6) 

0.0408 



8.0(16.8) 

25.7(21.8) 

0.0001 



0.1 (0.5) 
4.3(7.1) 
0.0001 



0.0 (0.0) 
2.4(9.0) 
0.0001 



15.4(20.8) 

29.0 (37.9) 

0.4724 



58.8 (33.3) 

48.7(37.4) 

0.0310 



74.4 (28.6) 

66.8 (34.4) 

0.0072 



Table 6-5— continued. 



614 



Parameter 



Ludwigia 

alala 



Ilea 
virginica 



Smilux 
waheri 



Smilax 
laurifolia 



Sample Size, Species Present 
Sample Size, Species Absent 

■ Water Depth (SD) (m). Present 
I Water Depth (SD) ( m). Absent 
P>t 



■ % Time in DC1 (SD). Present 
I % Time in DC 1 (SD). Absent 
P>Z 



x % Time in DC2 (SD). Present 
» % Time in DC2 (SD), Absent 
P>Z 



x % Time in DC3 (SD), Present 
1 % Time in DC3(SD), Absent 
P>Z 



x % Time in DC4 (SD), Present 
x % Time in DC4 (SD), Absent 
P>Z 



• % Time in DCS (SD), Present 
J "/.Time in DC5(SD), Absent 
P>Z 



x % Time in DC6(SD). Present 
I °/o Time in DC6(SD). Absent 
P>Zt 



i % Time in DC7(SD). Present 
R % Time in DC7 (SD), Absent 
P>Z 



x Overetory % Cover (SD), Present 
x Overetory % Cover (SD), Absent 
P>Z 



II % Low Level Light Available (SD). Present 
x % Low Level Light Available (SD). Absent 
P>Z 



x % High Level Light Available (SD). Present 
. % High Level Light Available (SD). Absent 
P>Z 



6 
938 

0.29 (0.09) 

0.20(0.21) 

0.2767 



29.0 (6.2) 

20.0 (22.2) 

0.0394 



3.6(0.4) 
5.7(4.8) 
0.4708 



7.4 (0.3) 

16.6(13.4) 

0.0491 



11.9(0.7) 

26.5(16.1) 

0.0181 



20.0(1.5) 

24.9 (22.0) 

0.9838 



18.2(2.8) 
4.0 (7.0) 
0.0002 



9.9(3.0) 
2.3 (8.8) 
0.0001 



16.5 (33.9) 

28.4 (37.4) 

0.3977 



57.7(23.8) 

49.1(37.4) 

0.6619 



76.0 (9.8) 

67.3 (34.3) 

0.5568 



95 
828 

0.18(0.27) 

0.20 (0.20) 

0.5085 



26.6(25.2) 

18.9(21.2) 

0001 



6.5(4.3) 
5.5 (4.7) 
0.0068 



16.5(10.8) 

16.5 (13.7) 

0.1623 



21.4(11.8) 

27.0(16.3) 

0.0067 



18.5(14.2) 

25.9(22.5) 

0.0200 



6.5(10.0) 
3.9(6.6) 
0.0010 



4.0(10.4) 
2.2 (8.6) 
0.1555 



47.3 (38.2) 

25.7(36.4) 

0.0001 



33.9(28.9) 

50.8 (37.7) 

0.0001 



48.3(33.1) 

69.6 (33.5) 

0.0001 



38 

451 

0.06(0.19) 

0.19(0.16) 

< 0.0001 



34.7(29.2) 

19.5 (20.4) 

0.0001 



7.7(4.3) 
5.6(4.5) 
0.0012 



17.9 (10.4) 

17.2(13.4) 

0.1199 



22.4(11.8) 

27.1 (14.9) 

0.0986 



15.4(14.3) 

24.8(21.1) 

0.0073 



1.6(2.3) 
4.0 (5.9) 
0.5023 



0.3 (0.9) 
1.9(5.4) 
0.0762 



60.2 (37.5) 

27.9 (37.7) 

0.0001 



27.8 (28.9) 

50.4 (37.6) 

0.0009 



37.3 (36.0) 

67.9 (33.9) 

0.0001 



49 
440 

0.09(0.14) 

0.19(0.17) 

0.0001 



25.0 (24.7) 

20.2(21.1) 

0.1116 



10.2 (6.3) 
5.2 (4.0) 
0.0001 



31.7(18.1) 

15.6(11.4) 

0.0001 



25.0(16.2) 

26.9(14.5) 

0.1341 



5.4 (6.9) 

26.1 (20.8) 

0.0001 



1.4(4.7) 
4.0(5.8) 
0.0001 



1.3 (5.2) 
1.8(5.2) 
0.0012 



47.9(40.2) 

28.5 (38.0) 

0.0004 



33.7 (34.7) 

50.3 (37.4) 

0.0139 



53.6(38.0) 

66.9(34.4) 

0.0354 





Table 6-5~continued. 








615 




Parameter 


Cepkalantkus 


Clethra 


Cyrilla 


Pierus 








occidental* 


alnifolia 


racemiflora 


pkiltyreifolia 




Sample Size, Species Present 


15 


12 


119 


34 




Sample Size. Species Absent 


474 


477 


370 


455 






X Water Depth (SD) (mX Present 


0.21 (0.19) 


0.20(0.16) 


0.09(0.12) 


0.11(0.15) 






J Water Depth (SD) (m). Absent 


0.18(0.17) 


0.18(0.17) 


0.21(0.17) 


0.18(0.17) 






P>t 


0.4011 


0.7048 


0.0001 


0.0170 






2 % Time in DC 1 (SD). Present* 


31.1 (15.0) 


15.3(11.0) 


26.6(21.4) 


26.5 (23.2) 






x % Time in DC1 (SD). Absent 


20.4(21.6) 


20.8(21.7) 


18.8(21.3) 


20.3 (21.4) 






P>Z 


0.0023 


0.8604 


0.0001 


0.0470 






'a % Time in DC2 (SD). Present 


4.5 (2.8) 


9.0 (6.4) 


9.0 (4.7) 


9.9 (6.2) 






X STime in DC2 (SD). Absent 


5.8(4.5) 


5.7 (4.4) 


4.7(3.9) 


5.4(4.2) 






P>Z 


0.4708 


0.0400 


0.0001 


0.0001 






x % Time in DC3 (SD). Present 


9.9 (6.3) 


26.2 (15.4) 


26.0(13.5) 


24.6(14.4) 






« % Time in DC3 (SD). Absent 


17.4(13.2) 


17.0(13.0) 


14.4(11.7) 


16.7(12.9) 






P>Z 


0.0152 


0.0155 


0.0001 


0001 






s % Time in DC4(SDX Present 


13.7(7.3) 


28.3(11.8) 


26.7(13.6) 


25.6(14.8) 






5 % Time in DC4(SDX Absent 


27.1 (14.7) 


26.7(14.8) 


26.7(15.1) 


26.8(14.7) 






P>Z 


0.0001 


0.6935 


0.5843 


0.3374 






> % Time in DCS (SD). Present 


18.8 (5.9) 


16.0(15.9) 


10.0(10.0) 


11.2(13.1) 






x %Time in DC5 (SDX Absent 


24.2(21.1) 


24.2 (20.9) 


28.5(21.4) 


25.0(21.0) 






P>Z 


0.7950 


0.1735 


0.0001 


0.0001 






8 % Time in DC6 (SD). Present 


13.7(7.4) 


2.5 (4.8) 


1.0(2.3) 


1.3(3.1) 






'■ % Time in DC6 (SD). Absent 


3.5(5.4) 


3.8(5.8) 


4.7(6.2) 


4.0 (5.9) 






P>Zl 


0.0001 


04850 


0.0001 


0.0094 






! % Time in DC7(SD). Present 


8.3(5.8) 


2.7(9.3) 


0.6 (3.3) 


1.0(5.5) 






! % Time in DC7(SD), Absent 


16(5.1) 


1.8(5.1) 


2.2 (5.7) 


1.9(5.2) 






P>Z 


0.0001 


05268 


0.0001 


0.0061 






1 Overstory% Cover (SDX Present 


38.8(49.2) 


54.3 (38.5) 


45.4(39.5) 


62.1(38.0) 






x Overstory % Cover (SDX Absent 


30.2 (38.2) 


29.8 (38.4) 


25.6(37.1) 


28.1 (37.6) 






P>Z 


0.9257 


0.0275 


0.0001 


0.0001 






s % Low Level Light Available (SDX Present 


38.7(35.9) 


29.1 (32.9) 


31.8(31.9) 


18.6(24.3) 






x % Low Level Light Available (SDX Absent 


48.9 (37.5) 


49.1 (37.5) 


54.0 (37.6) 


50.8(37.3) 






P>Z 


0.2215 


0.1688 


0.0001 


0.0001 






J % High Level Light Available (SDX Present ' 


53.7(32.9) 


41.6(37.7) 


52.7(37.4) 


37.9(35.9) 






x % High Level Light Available (SDX Absent 


65.9(35.0) 


66.2 (34.7) 


69.7 (37.4) 


67.6(34.1) 






P>Z 


0.0130 


0.0352 


0.0001 


0.0001 
















J 



Table 6-5-continued. 



616 



Parameter 



1 yon in 
luclda 



Leucothoe 
racemosa 



Gordonia 
lasUmtkus 



Ilex 
cassine 



Sample Size. Species Present 
Sample Size. Species Absent 

; Water Depth (SD) (m). Present 
a Water Depth (SD) (m). Absent 
P>t 



S % Time in DC 1 (SD). Present 
« % Time in DC1 (SD). Absent 
P>Z 



a % Time in DC2(SD), Present 
a % Time in DC2 (SD). Absent 
P>Z 



5 % Time in DC3 (SDX Present 
x % Time in DC3 (SD). Absent 
P>Z 



J % Time in DC4 (SD), Present 
« % Time in DC4 (SD), Absent 
P>Z 



a % Time in DC5 (SD), Present 
5 %Time in DC5 (SD). Absent 
P>Z 



a % Time in DC6 (SD), Present 
I % Time in DC6 (SD), Absent 
P>Zl 



5 % Time in DC7(SD), Present 
a % Time in DC7 (SD), Absent 
P^Z 



* Overstory% Cover (SD). Present 
a Overstory % Cover (SD), Absent 
P>Z 



a % Low Level Light Available (SD), Present 
a % Low Level Light Available (SD), Absent 
P>Z 



J % High Level Light Available (SD), Present 
a % High Level Light Available (SDX Absent 
P>Z 



93 
396 

0.07(0.12) 

0.20(0.17) 

0.0001 



28.1 (24.3) 

19.0 (20.5) 

0.0001 



10.1(5.8) 
4.7 (3.4) 
0.0001 



27.3(14.7) 

14.8(11.5) 

0.0001 



26.2(15.4) 

26.9(15.4) 

0.1486 



7.8 (9.3) 

27.9(20.9) 

0.0001 



4(0.9) 
4.6(6.1) 
0.0001 



0.0 (0.0) 
2.2(5.7) 
0.0001 



61.7(36.9) 

23.1(35.2) 

0.0001 



22.0 (28.8) 

54.8 (36.6) 

0.0001 



39 4 (36.6) 

71.7(31.7) 

0.0001 



68 

421 

0.07(0.11) 

0.19(0.17) 

0.0001 



28.1 (23.6) 

19.5(21.0) 

0.0005 



9.5 (3.9) 
5.1(4.3) 
0.0001 



28.9(15.0) 

15.3(11.8) 

0.0001 



24.8(11.7) 

27.0(15.1) 

0.2588 



7.0 (6.6) 

26.8(21.0) 

0.0001 



1.1 (3.4) 

4.2 (5.9) 
0.0001 



0.5 (2.3) 
2.0 (5.5) 
0.0001 



54.0 (39.2) 

26.6 (37.2) 

0.0001 



29.2(31.9) 

51.7(37.4) 

0.0001 



43.8 (37.9) 

69.1 (33.2) 

0.0001 



33 
133 

08(0.10) 

0.20(0.16) 

0.0001 



24.9(21.8) 

18.4(18.6) 

0.2010 



9.8(5.0) 
5.2 (3.8) 
0.0001 



28.0(13.5) 

14.9(10.3) 

0.0001 



27.2(13.1) 

26.6(13.0) 

0.9419 



7.9(11.2) 

27.9(19.9) 

0.0001 



0.2(0.5) 
4.4 (5.9) 
0.0001 



0.0(0.0) 
2.1(5.6) 
0.0005 



57 
109 

0.12(0.16) 

0.20(0.15) 

0.0020 



24.4 (20.8) 

17.3(18.2) 

0.0048 



8.1(4.9) 
5.0(3.8) 
0.0001 



20.7(11.8) 

15.8(12.0) 

0.0006 



25.8(13.0) 

27.2(13.0) 

0.3914 



15.6(14.4) 

28.3(21.4) 

0.0001 



2.5 (5.0) 
4.2(5.7) 
0.0142 



1.9(5.3) 
1.2(4.6) 
0.0117 



Table 6-5-continued. 



617 



Parameter 


Magnolia 


Persea 


Pints 


Nyssa 




virginiana 


palustris 


spp. 


sylvatica 
v. biflora 


Sample Size. Species Present 
Sample Size, Species .Absent 


20 
146 


7 
159 


12 
154 


34 
132 


s Water Depth (SD) (m). Present 
; Water Depth (SD) (m). Absent 
P>t 


0.11(0.13) 

0.19(0.16) 

0.0572 


0.11(0.09) 

0.18(0.16) 

0.2928 


0.15(0.11) 
0.18(0.16) 

0.5773 


0.22 (0.21) 

0.17(0.21) 

0.1549 


II % Time in DC1 (SD), Present 
II % Time in DC1 (SD). Absent 
P>Z 


26.9(19.6) 

18.7(19.2) 

0.0356 


25.6 (19.3) 

19.4(19.4) 

0.2589 


13.5(16.0) 

20.2(19.6) 
0.1625 


26.5(18.6) 

17.9 (19.2) 

0.0020 


■ % Time in DC2 (SD), Present 
1 % Time in DC2 (SD), Absent 
P>Z 


9.7(5.2) 
5.6(4.1) 
0.0002 


8.0(2.9) 
6.0(4.5) 
0.0819 


11.0(8.1) 
5.7(3.8) 
0.0103 


4.3(3.2) 
6.5(4.6) 
0.0059 


n %Time in DC3 (SD). Present 
x % Time in DC3 (SD), Absent 
P>Z 


23.0(12.1) 

16.7(12.0) 

0.0127 


20.5 (3.6) 

17.3(12.4) 

0.0908 


27.0(15.5) 

16.7(11.6) 

0.0109 


10.4(6.4) 

19.3 (12.6) 

0.0001 


1 % Time in DC4(SD). Present 
5 % Time in DC4(SD), Absent 
P>Z 


24.4(9.8) 

27.0(13.3) 

0.4273 


30.1 (15.3) 

26.5(12.9) 

0.8002 


30.1(13.1) 

26.4(13.0) 

0.3577 


18.2(11.3) 

28.9(12.4) 

0.0001 


7 % Time in DC5 (SD). Present 
x % Time in DC5 (SD), Absent 
P>Z 


12.8(13.2) 

25.4 (20.5) 

0.0042 


14.8(7-3) 

24.3 (20.5) 

0.3682 


18.2 (29.9) 

24.4(19.3) 

0.0254 


22.3(11.8) 

24.3 (21.8) 

0.5470 


1 % Time in DC6(SD). Present 
x % Time in DC6(SD). Absent 
P>Zt 


2.3 (5.0) 
3.8(5.0) 

0.1678 


1.0(0.6) 
3.7(5.6) 
0.6145 


0.0 (0.0) 
3.9(5.6) 

0.0001 


10.3 (7.6) 
1.8(3.0) 
0.0001 


> % Time in DC7 (SD), Present 
1 % Time in DC7 (SD). Absent 
P>Z 


9(3.5) 
1.8(5.3) 
0.4459 


0.0(0.0) 
1.7(5.2) 
0.2620 


(0.0) 
1.8(5.2) 
0.0197 


7.4(8.9) 
0.2(1.4) 
0.0001 


x Ovemory % Cover (SD). Present 
x Overstory % Cover (SD), Absent 
P>Z 


d 


d 


d 


d 












Table 6-5--continued. 



618 



Parameter 



Acer 

rubrum 



Taxodium 
ascendens 



Nyssa 
ogeechee 



Ilex 

myrtifolia 



Sample Size. Species Present 
Sample Size. Species Absent 

; Water Depth (SD) (m). Present 
II Water Depth (SD) (m). Absent 
P>t 



8 
158 

0.08(0.19) 

0.18(0.16) 

0.0873 



1 •/. Time in DC 1 (SD), Present 
1 % Time in DC 1 (SD), Absent 
P>Z 


40.0(17.9) 

18.7(18.9) 

0.0029 


x % Time in DC2(SD). Present 
! % Time in DC2 (SD). Absent 
P>Z 


4.4 (3.3) 
6.2 (4.5) 
0.2471 


x % Time in DC3(SD). Present 
1 HTime in DC3(SD). Absent 
P>Z 


9.8(7.6) 
17.9(12.2) 

0.0371 


1 % Time in DC4(SD), Present 
9 % Time in DC4 (SD), Absent 
P>Z 


13.6(10.7) 

27.4 (12.7) 

0.0028 


x % Time in DCS (SD), Present 
t % Time in DCS (SD). Absent 
P>Z 


15.4(6.4) 

24.3 (20.5) 

0.3838 


x % Time in DC6(SD), Present 
» % Time in DC6 (SD). Absent 
P>Zt 


9.7(6.9) 
3.3(5.3) 
0.0033 


X % Time in DC7(SD), Present 
x % Time in DC7 (SD), Absent 
P>Z 


7.2 (5.9) 
1.4(4.9) 
0.0006 


x Overnory /. Cover (SD), Present 
; Overstory % Cover (SD). Absent 
P>Z 


d 



93 

73 

0.17(0.16) 

0.18(0.13) 

0.5789 



22.2 (20.5) 

16.5 (17.5) 

0.0338 



5.9(4.1) 
6.3 (4.8) 
0.6324 



15.5(10.1) 

20.0(14.0) 

1392 



25.5(13.1) 

28.2(12.7) 

0.2956 



23.0 (17.8) 

25.1(22.9) 

0.8785 



4.3 (5.9) 
2.7(4.9) 
0.0030 



2.6 (6.4) 
0.5(2.2) 
0.0363 



6 

160 

0.10(0.21) 

0.18(0.16) 

0.2124 



42.9(14.3) 

18.8(19.0) 

0.0029 



3.0 (0.8) 
6.2(4.5) 
0.0390 



6.0 (2.0) 

17.9(12.2) 

0.0025 



8.7(3.0) 

27.4(12.7) 

0.0004 



15.5(5.0) 

24.2 (20.4) 

0.4861 



13.9(4.2) 
3.2 (5.2) 
0.0003 



10.1(2.3) 
1.3(4.9) 
00O01 



3 
163 

0.09 (0.33) 

0.18(0.16) 

0.6811 



43.8(21.2) 

19.3(19.1) 

0.0495 



2.6 (0.5) 
6.2 (4.5) 
0.0788 



5.1 (1.6) 

17.7(12.1) 

0.0190 



7.6(2.7) 

27.0(12.8) 

0.0081 



14.5(6.1) 

24.1 (20.3) 

0.5444 



14.2(7.1) 
3.4(5.3) 
0.0116 



12.2(3.8) 
1.5(4.9) 
0.0010 



Table 6-5-continued. 



619 



Parameter 


Fraxinns 




caroliniana 


Sample .Size. Species Present 
Sample Size, Species Absent 


3 
163 


1 Water Depth (SD) (m). Present 
» Water Depth (SD) (m), Absent 
P>t 


0.26 (0.28) 

0.17(0.16) 

0.3607 


1 % Time in DC1 (SD). Present 
i! %Time in DC1 (SD). Absent 
P>Z 


33.7(21.9) 

19.4(19.3) 

0.1526 


7 % Time in DC2 (SD), Present 
« % Time in DC2 (SD), Absent 
P>Z 


5.0(3.1) 
6.1(4.5) 
0.7297 


J % Time in DC3(SD), Present 
s HTime in DC3 (SD). Absent 
P>Z 


9.7(4.4) 

17.6(12.2) 

0.1824 


i! % Time in DC4 (SD). Present 
S! % Time in DC4(SD), Absent 
P>Z 


11.5(1.2) 

27.0(12.9) 
0.0300 


1 % Time in DC5(SD), Present 
1 % Time in DC5(SD), Absent 
P>Z 


15.6(9.7) 

24.1(20.3) 

0.6278 


5 % Time in DC6(SD). Present 
S! % Time in DC6(SD). Absent 
P>Zt 


14.5(12.3) 
3.4(5.2) 
0.0980 


S % Time in DC7(SD), Present 
? % Time in DC7 (SD), Absent 
P>Z 


9.9(8.6) 
1.5 (4.9) 
0.0537 


R Overatory % Cover (SD). Present 
t Overstory % Cover (SD), Absent 
P>Z 


d 



Table 6-5-continued, 
a Depth Classes (DC) are: 



620 



(no inundation) DC 1 
(shallow water) DC 2 
(shallow water) DC 3 
(shallow water) DC 4 
(deep water) DC 5 
(deep water) DC 6 
(deep water) DC 7 



water depth < 0.0 m 
0.00 m < water depth < 0.05 m 
0.05 m < water depth < 0. 15 m 
0. 15 m< water depth < 0.30 m 
0.30 m < water depth < 0.60 m 
0.60 m < water depth < 1 .00 m 
water depth >1.00 m 



b Light availability measured at 0.3 m above the peat surface with a Licor quantum 
sensor. See chapter text for details. 

c Light availability measured at 1.0 m above the peat surface with a Licor quantum 
sensor. See chapter text for details. 

d Overstory % cover and availability of light at 0.3 m and 1.0 m above the peat surface 
were not estimated for tree belt transects. 






621 



Carex waiter ian a 

(n-44 of 437 plots) 



1 - 25 % Frequency quartiles of range of flooding 
E3 so 7s % duration where species occurrence was 
•3 75-100% Breatest(-n). 



■n). 

Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs, 
Maximum duration of flooding at sites sampled. 



--. >0.60-1.00 



f 

Q 



>o.3o-o.eo 



! 




>0.16-0.30 _ 



>0.05-0.16_ 



t — i — i — i — i — i — r — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Nymphaea odorata 

(n-50 of 361 plots) 



^ >0 .60-1 .00 

E 




[3^&&&^&%l 



-►• 







m%& * 



-►♦ 



-► • 



~~ i i i i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7. Hydrologic conditions where species occurred at greatest abundance (90- 
100% maximum density or percent cover). 



622 



Rhynchospora 
inundate 

(n-7 of 35 plots) 



□ 1 ." 2 ft, Frequency quart les of range of flooding 
= n ,r ., duration whore species occurrence was 
E3 75-100% 9™atB8t(-n). 



Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs 
Maximum duration of flooding at sites sampled. 




>o-0.06 
<-0 



-i^fffflfiftftfflft-^-y 



10 



r 

20 



I 

40 



T 



30 40 50 60 70 80 
% of Time in Depth Class 



i — r 

90 100 



Panicum hemitomon 

(n«26 of 228 plots) 



>1.00 



►• 




i i i i — i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7-continued 



623 



Lacnanthes 
caroliniana 

(n=29 of 266 plots) 



SO - 76 % 



Frequency quartiles of range of flooding 
duration where species occurrence was 



■ 

76-100% greatest (»n) 

O 



Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs. 

Maximum duration of flooding at sites sampled. 



.-■ >0 .60-1 .00 
^ >0. 30-060 

I 







Oi >0.15-0.30 - . 



I 



B- 



-►• 



ffi^&&&t- 



~ i i — i — i — i — i — I — i — r 

10 20 30 40 SO 60 70 80 90 100 
% of Time in Depth Class 



Eleocharis baldwinii/vivipara 

(n-25 of 157 plots) 



>1.00 



—» >0.60-' 



aassssasa 



a 

o 
Q 

O) 

c 

! 



>0.30-0.60 



10 o.^M— ► 

poo i mmm 




10 20 30 40 SO 60 70 80 90 100 
% of Time in Depth Class 



Fjgyn? 6-7-continved. 



624 



Nuphar luteum 

(n=1 2 of 68 plots) 



E >0. 60-1.0 






1-26% 

25 • 50 % 

60-75% 



■ 

GS 75-1 

o 



Frequency quart les of range of flooding 
duration where species occurrence was 
greatest (- n). 



Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs, 

Maximum duration of flooding at sites sampled. 



a 
a> 

O 

9 



X>. 30-0.80- 



■H >0.06-0.1 



BB 




10 20 



40 50 60 70 80 90 100 
% of Time in Depth Class 



Eriocaulon lineare 

(n«7 of 44 plots) 



£ >0 .60-1. 00- 
S" >0 .30-0.60 

O 

* >0.16-0.30- 

1 

2. >0.06-0.15- 

E 

>O-O05 
<»0 _ 



-►• 



m- 



■*► • 



_+t%M— ► • 



£n* 



n — i — i — i — i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 






Figure 6-7-continued 






625 



Xyris species 

(n-26 of 248 plots) 



3 »s Z fiott Fr »quency quartilesof range of flooding 
»*•••< duration where species occurrence was 

S£? 50-70% t_i# & 

E3 7S-ioo% 8™atest(=n). 

O Mode of flooding duration where species 

occurrence was greatest 

■*♦■ Range of flooding duration where species occurs. 
• Maximum duration of flooding at sites sampled. 




10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



>1.00 



■=• >0 .60-1 00 



f 

e 

o 
a 

c 



>0.30-0.60 



>0.15-0.30- 



>0 .06-0.16- 



> 0-0. 06 
<*0 



Peltandra virginica 

(n=78 of 356 plots) 




sssssss^ssffii^ 



^a^assssssaa* 



\ 






-►• 



10 



T 

20 



T - 

60 



I 
70 



T 



30 40 50 60 70 80 
% of Time in Depth Class 



90 100 






Figure 6-7-continued 









626 



Sphagnum species 

(n=34 of 308 plots) 



1-25% Frequency quartlles of range of flooding 
E3 er. , c .' duration where species occurrence was 
ES 76-100% greatest (-n). 

O Mode of flooding duration where species 

occurrence was greatest. 

Range of flooding duration where species occurs 

Maximum duration of flooding at sites sampled. 



>1 00 tft.^.t.^.i.^.t.i.t.t.Kt.iVtK.t.i.tK 1 



«..«..i.i..«..v.«..i..*..^..i.....i...........o...«..>.-....o«..o«; 



3>* 308 



■=• >0. 60- 1.00. 

^. >0.3C 

I 

a 

2> x>i« 



>0-0.06 



A 



S-S-WS-S-\-S-..S.%-.sW 




m- 



-►• 



■■■.^.^.■■■.■■•■■^■■.■^■.■..■■■■■i.^....^. 



3-** 



-►• 



^MMM^MM^ - 



10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Utricularia species 



(n«35 of 244 plots) 

,,.00 |ktftftftftftfSatftftftftftft3 




60 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7-continued 



627 



Woodwardia 
virginica 

(n=29 of 197 plots) 



>i.<x> . r 



>0.S0-1.00. 



5 
a 
a> 
a 

CO 

■o 

§ 



>0.30-0.60. 



m YB\i\iVi%i\i%i\ 



X).06-0.15-" 



3 m 'snt Frequency quartiles of range of flooding 

E3 •>*•?•. duration where species occurrence was 

*** DO - #5 % . . . . 

Q 75-100% 9™atest(«n). 



Mode of flooding duration where species 
occurrence was greatest. 

Range of flooding duration where species occurs. 

Maximum duration of flooding at sites sampled. 



-*• 



WMWMmm>?%8gL 



-** 



^&&&&&B&&i . 






-►• 



...i 






T 

10 



20 



I 

40 



1^ 
70 



30 40 SO 60 70 80 
% of Time in Depth Class 



i r 

90 100 



Itea virginica 

(n-10of95plots) 







R3ss%^?^^&a&&ae 



~i i I i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 



% of Time in Depth Class 



Figure 6-7-continued 



628 



Lyonia lucida 

(n-10of93plots) 

i 



■~ >o.eo-i.oo_. 



a 
o 

a 



a 

c 

\ 

o 

a. 



>0.15-0.30_. 



JO >0 05-0.15 _. 



□ L 2 f«t Frequency quartiles of range of flooding 

*~ i 25 - 60 % 



B 



duration where species occurrence was 



E3 76-100% 0">a»»*M«n) 

O Mode of flooding duration where species 

occurrence was greatest. 

Range of flooding duration where species occurs. 

Maximum duration of flooding at sites sampled. 




a^&&&&&& 



w&+ 



-> • 



i 

40 



60 



30 40 50 60 70 80 
% of Time in Depth Class 



i — r 

90 100 



-^ >0 .80-1 .00 






■{£ X) .30-0.80 _ I 

I 
D 

0> >0 15-0.30 _ 

T> 
O 

,0 >0 05-0 15- 
LL 



>0-0.06 
<»0 



fcll- 



Leucothoe 
racemosa 

(n»7 of 68 plots) 



^^- 



i — i — r 



i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7-continue^ 



629 



Smilax lauri folia 

(n«6 of 49 plots) 



CI L" 2 f* Frequency quartiles of range of flooding 

Sd 25-60% _j.._^ai__ „^____ : 

_ 60-75% 

Q 75-100% greatest (-n). 



Mode of flooding duration where species 
occurrence was greatest. 

Range of flooding duration where species occurs, 
Maximum duration of flooding at sites sampled. 




"0 <l^vV;V-.^s^^^s^^^s^^-.tfs^T- 



-►• 



~i — i — i — i — i — i — i — i — i — r 

10 20 30 40 SO 60 70 80 90 100 
% of Time in Depth Class 



^» >0.60- 



a 



Smilax waiter/ 

(n=5 of 38 plots) 



a 

c 

J 

o 



>0.16-0.30 




% of Time in Depth Class 



Figure 6-7-continued. 



630 



Magnolia virginiana 

(n=4 of 20 plots) 



□ \" 2 f |* Frequency quartiles of range of flooding 
E] * n -»«-*, duration where species occurrence was 
75- 100 S Breatest(-n). 



O Mode of flooding duration where species 

occurrence was greatest 

♦► Range of flooding duration where species occurs 
• Maximum duration of flooding at sites sampled. 




% of Time in Depth Class 



Pierus phillyreifolia 

(n*4 of 33 plots) 



« ► 



£ >0 30-0 I 

e 
O 

O) >0 15-0 30 _ 

c 

■D 

8 >0.06-0.' 
U. 



A 



>o-o.os 

<-0 



4 



gsg^ 



.*-£ 



■>• 



-»-• 



-► • 



T 



i i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7~coptjrme4 



631 



Cordon ta /as/a nth us 

(n«4 of 40 plots) 



13 as hm Frequency qualities of range of flooding 

E3 ™ -, duration where species occurrence was 

SB 60 - 76 % . *_ v 

H 76-100% oreatest(-n). 



Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs 
Maximum duration of flooding at sites sampled. 




% of Time in Depth Class 



Taxodium ascendens 

(n-10 of 190 plots) 



-=■ >0 .60-1 00 
^. >0 30-0.60 



w m$$m 




i 



g >0 15-0 30. 

O 

O >0.05-0.16_ 



■rf wmm — 



>0-0.05 



«0 



\WL, 



*&&&&& 



~i — i — i — i — i — i — i — i — i — r 

10 20 30 40 SO 60 70 60 90 100 
% of Time in Depth Class 



Figure 6-7-continued 



632 



Ilex cassine 

(n"6 of 57 plots) 



□ 2s 2 5o% Frequency quartilea of range of flooding 
3„'«„ duration where species occurrence was 
E 76-100% greatest («n). 



Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs. 

Maximum duration of flooding at sites sampled. 



? 

a 
o 


>0 60-1.00 




>0 .30-0.60 


0) 

r 


>0 15-0.30 


8 




>0 .06-0.1 S 




'0 _* 



t — r 

40 SI 
% of Time in Depth Class 



Nyssasylvaticav. biflora 

(n"4 of 34 plots) 




t i — i — i — i — i — r 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7-continued 



633 



Frequency qualities of range of flooding 



S 1 ■ 2S % 

E3 en tk v duration where species occurrence was 
Q 76-100% flreatest(=n). 



Mode of flooding duration where species 
occurrence was greatest 

Range of flooding duration where species occurs 
Maximum duration of flooding at sites sampled. 



>0 30-060 



a 
a 

a 

a 

_c 

■D 
O 
_0 
u. 



>01S-0.30 _ 



>0 .05-0.15 



>0-0.05 _ 



Cyril la racemiflora 

(n=13 of 119 plots) 



>1.00 _ i » 

>0 .80-1 .00 _ iWftftfoj ^ 



^^^mm^m . 




>. i mmm^M^MMM^i 



10 



20 



T 

30 



T 

40 



50 



I 
60 



► • 



T" 

70 



% of Time in Depth Class 



n — i — r 

80 90 100 



Figure 6-7-continued. 



634 



Understory Species 
Diversity 

(n=93 of 937 plots) 



□ 
E2 



i - 25 % Frequency quarti les of range of flooding 
■?*«■ 2 duration where species diversity was 

*** 50 - 75 % . . . . 

ES 76-100% greatest (- n). 

O Mode of flooding duration where species 

diversity was greatest. 

Range of flooding duration where species occur. 
Maximum duration of flooding at sites sampled. 



>1.00 



J 



>0 .60-1 .00. 



£ 

0- 
O 

O 
o> 

c 

1 



O >0 .05-0.1 5 



-*+ 



sssssssssssz - 



-*» 






>0. 15-0. 30 



>0-0.06 




° ) rnxxxsMss®^^ 



I 

10 



20 



T 

40 



50 



I 

70 



-►• 



30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Shrub Species 
Diversity 

(n»16 of 157 plots) 




T I I — I — I — I — I — I — T 

10 20 30 40 50 60 70 80 90 100 
% of Time in Depth Class 



Figure 6-7~continued. 



635 



I 26-60% frequency quartites of range of flooding 
3 60 75% duration where species diversity was 
Q 75-100% 9reatest(«n). 

O Mode of flooding duration where species 

diversity was greatest. 

'* Range of flooding duration where species occur. 
• Maximum duration of flooding at sites sampled. 



Tree Species 
Diversity 

(n*52of84plots) 



' fe&&&&?^ 



£ 

c 
•o 
§ 



>o.eo-i.oo 



>0.30-060 



>0.16-0.30 



>0 .05-0.1 5 



\z>7X 



msm 



WMg^MMMMMS^ 







I — i — i — i — r 

60 70 80 90 100 



% of Time in Depth Class 






Figure 6-7~continnr^ 



636 
associations could be identified by examining similar species' distributions relative to 
flooding extent and duration, and light availability. 

Sufficient data were available to calculate regression models for 26 species. 
Initial efforts focused on modeling data from all samples (n=942) for all 26 species, 
regardless of species presence. When considering all sample sites, most herbaceous 
species were negatively associated with all the modeled hydrologic variables, suggesting 
that their infrequent occurrence throughout the swamp resulted in a significant departure 
from zero whenever they did occur (Table 6-6). Therefore, these species were not further 
examined with the "all sample" models, but were further analyzed with the "species 
present" models, which would elucidate the species-environment relationships where the 
species occurred. Abundances of several shrub and tree species were significantly 
related to selected hydrologic parameters in the "all sample" models, and, unlike the 
herbaceous species, single depth classes were responsible for this relationship when all 
samples were considered. None of these species were significantly related to inundation 
duration in deep water (depth > 0.30 m). Trees species were generally associated with no 
inundation, whereas shrub species were inundated to shallow depths (Table 6-6). Light 
availability was not significantly related to occurrence for most shrub species, although a 
few were in greater abundance where light availability was low; most likely this low light 
availability was a function of the shrubs themselves, which generally were between 1-2 
m in height. For those species significantly associated with shallow water depths (0-0.30 
m), species' densities were most often correlated with duration of flooding to 0.05-0.15 
m (Table 6-7). Tree and shrub diversities were greatest with inundation depth of 



637 



Q. 

E 

a 
v> 

8 



•a 

c 

CO 
Vi 

3 
O 
<D 
O 

a 
| 



CO 

M 

c 

3 

VI 

U 
1 



o 








VI 




(/I 




<i* 




&) 




a; 




V- 




<D 








D. 








•*■■» 








3 




£ 




•*-» 




c 




<D 




E 




c 




2 


u 




o 


> 


c 


c 


« 


<u 


VI 


1 

VI 


o 


<L> 


Q. 


u 


vi 


6, 

V> 


1) 


<u 


i 


x: 


v> 




'— 


c 









f= 


VI 
VI 

1) 

i_ 


rt 


ftij 


■— 


a> 


<rt 


u. 


O. 


/ -^ s 




T3 


c 


0> 


A 


c 


O 


J— 


c 


,P 




-t- 


& 


C 




(Tt 


r/") 


l- 




■*-» 






i 


00 


VO 


o 






a; 


■_- 


X) 


ca 


« 


9 


H 


•a 



u, 

A 



B-tl! 



i-ISl 



ill-ii 



f? Ill 



B"lll 



8"IJ1 



it-ji 



8-ii 



15 



III 



« I t 

1-5 



Hi 



g-8 



g-g 



i 

i 



M M *S IN 



* i 



i j 



* ! 



I t J i t i 



t- 



i 



* i 






I 



SI li 



, 



i- 



|| 

II 



1 S 



1 I 









638 



u. 
A 
a. 



8-IS! 



u-f 



a H fsf 



8 8 



8 8 8 



!H»Jl 



5!S 



T3 

U 
3 
S 

c 
o 

o 

I 

v© 

•a 



B-I!l 



M!i 



Mi 



if 



S- ■ 



515 



i» 



SIS 



8-8 



B"B 



2- 



i t t 



i i i 



* i i 



ii i 



ill 



I! 



li 



&3 



gi- 



ll ii 
1 1 If 



11 



639 





c 


■ 


! 


8 Depth Classes (DC) are: no inundation DCl water depth < m 

shallow water DC2 < water depth < 0.30 m 
deep water DC3 water depth > 0.30 m 


b Cell entries represent slope and a-levels for tests of significance as follows. 

+++,--- P< 0.001 






fa 
A 


© 


d 






> 


d 


d 






B-tl! 








DCl 

X 

High 
Level 
Light 








!l!"Jl 








DC3 

X 

Low 
Level 
Light 
















DCl 
X 

Low 
Level 
Light 








jf-ji 








b-ji 








<1 


J 


+ 






111 


t 






ill 




— < «n o 
© © — 

d d d 




all 




VI VI VI 

G- c_ o. 




8-8 


1 1 

I 1 






8-8 


: t 


i 




8-8 




+ + + 


-o 

CJ 
3 

5 




B 


+ 






*** 


i 


i 




1 
o 

CJ 

i 

i 


& 


t 




so 

u 

1 


i 

■ 


Shrub 

Species 
Diversity 


w '2 

!l 

- Q 





O 






o. 

D 
| 

c 



.8 

■4-* 
— 

o 

u 
o 

c 

CJ 

to 
X> 

C3 

u. 

<2 

•o 
a> 

o 

E 

c 
u 

4J 
X) 
co 

co 

x: 



c 
5 



CJ 

o 
o 

c 
o 

to 

CO 

u 

L- 

V 

a. 



3 

E 

CJ 

xs 



o? 



640 



en 




4> 








O 




k 




■J-, 




>> 




TJ 




O 




O 




£ 




"O 




c 




cfl 




cfl 




3 




O 




CD 




o 




<s 




X 




h* 




<L) 


^^ 


X! 


E 






Cfl 


o 




ro 


c 


O 


c/i 


VI 


3 


X 


co 




<D 


« 


T3 


•o 


O 

B 


V 




o 


S 




o 


£ 


Cfl 

CO 


^o 


& 


CO 

X 


<D 


cfl 


u- 




i> 


<U 

— 


o. 


eo 




Cfl 




X 


3 

E 






"O 


c 


n 


E 


•** 


c 


cfl 


o 


TD 


k. 


C 


> 


3 


c 


U 


« 




1 


C 


cfl 

<D 


X 


u 


it 


1 

Cfl 




o 


c 


x 


u 


■4-* 


Cfl 


c 


g 


cfl 


a 


— 


t/i 


u 


« 


■*-» 




1) 


u 


E 

S 


21 

Cfl 


m 


— 


Q. 


o 


■•-» 


Cfl 


c 


cfl 


(fl 


<D 




1 


1 


So 


co 


c 

2 


t»> 


cfl 


VO 


a> 


u 


c 


X> 
cfl 

» 


E 

Cfl 



A 

A. 



HI 






I 



* 



a 
d 



I 3 



8-2 



gS 



8- 



■ 



t- f I 



U 
if 

3 2 






if 



E S 55 

ItJ 

3 cflS 



tig 

ill 

c« cfl Q 



P a 



O 



E 
2 E 

°°, 

VI o 

•S vi 

' a 
D 



o 

VI ^ 

1| 

cc 

V 

E 



41 

T3 



41 



VI 

o 



cfl 

V 



o — 
o o 



M n rf 

u y u 

DQQ 

i 

U 
O 

6 

■ 

■ 

u 

€ 

a. 
u 

u 
V 

Cfl 

I 











I 

1 

a 
s. 

1 

i 
i. 




| 










_o 






o 




"5 






8 




<*5 






c 




en 
CO 


o 


O 


y 

en 




U 


o 


o 


CO 




c 


VI 


VI 


Ih 

o 




CO 




c- 


CU 


(S 




IG 






T3 






I 


| 


1 


c 




en 
O 


+■" 


+" 


1 




2 

en 

8 






c 

1 




5 






1 




en 
















+s 




1) 


*_ 




c 




> 


o 


^■H 


.32 







o 


o 


o 




i 

a 


o 


© 


£ 




•a 


VI 


VI 


6 




c 

CO 


PL. 


Oh 


o 
o 




1 






c 

en 




c/5 


I 




en 





— 

c 


1 


1 


& 


"O 


D 




1 


cu 


g 


en 


* 




E 


3 
C 

1 


Q, 

e 


-r 
+ 
+ 


+ 


.a 


O 
o 

1 
1 


en 






1 


r- 


-4^ 






J 




c 
6 






u 


— 






t* 










*N 


5 


u 






B 



641 



642 
0.05-0.15 m, and shrub diversity was greatest when overstory cover was high; understory 
diversity was high regardless of inundation conditions or light availability (Table 6-6). 
The transect variable was significantly associated with several species, suggesting that 
species were not uniformly distributed throughout the areas (Table 6-7). 

Significant model parameters that describe environments where selected species 
were present are listed in Table 6-8 (flooding conditions are no inundation, shallow 
water, or deep water) and Table 6-9 (division of shallow water depth as 0-0.05 m, 0.05- 
0. 15 m, and 0. 15-0.30 m). Woody species were usually associated with shallower 
conditions than herbaceous species, and light levels and transect were significant 
parameters for less than half of the sampled species (Table 6-8). Species associations are 
grouped by average and standard deviation of water depths in Table 6-10. 

Inundation conditions and durations described in Table 6-7 by the species' 
multiple regression models are represented in 3-dimensional plots in Figure 6-8. These 
plots illustrate the observed data and regression model-predicted species abundances 
with inundation depth and duration. Light availability and transect parameters are 
included in the models where appropriate, but are not diagramed in these plots. 
Although the figure base is a square to facilitate viewing, the modeled surface is 
confined to the lower right half creating a surface triangle. Two of the three modeled 
inundation parameters are illustrated on the axes (% time with no inundation and % time 
in shallow water), and the third parameter (% time in deep water) is the difference of 
these parameters from 100% (100% -% time no inundation - % time in shallow water = 
% time in deep water). For example, hurrahbush (Lyonia lucida) was found in greatest 



643 



u 

EL 

E 

a> 



T3 
C 
CO 



CO 

o 



D 

C 
3 

SO 

a 

lo 

3 



o 

•o 

o 

E 

c 
_o 

(7) 
U 

6b 

0) 

ha 

EL 



3 

E 
•*^ 

c 

<D 

E 

c 
o 

'> 
c 

U 
■ 



| 



<L) 



C/5 

u. 
<U 



E 
2 

e 

CO 

o 

ta 
I 

1/5 



00 
I 

•s 



\— 
CO 
(A 

- 

I 

(A 

u 

0) 

— 

CO 

i 



b. 
A 
- 



n All 
2S M = 55 



Sll-ll 



■ 1 £ 

11! 



II-Ji 



Ml 



8"|| 



Jl 



- - - 

HI 



Bis 



111 



§-2 






- 

is. 



r 



p 

: 



E 

I 
.a 

I 



644 



- 

A 

0. 



i i 

© © 



s = 

6 © 



1) 

3 
,C 

c 
o 
o 

00 
I 

o 
o 

H 



UHi 



8 -J! 



6* 



<- b 

s s 

6' 



B"JI 



II 



6U 



£2% 



111 



S-8 



S 



I 



I ! 



I I 



i 



| 



| 



I 



I 



s 2 

If 



II 

II 

a: E 



1 

I 



I 



ft. 



t 

I 

i 



u 
3 
C 



oo 
1 



c 


r- «n 

•r-V *© 


lb 

A 

0. 


i 1 

© © 
© © 


> 


© © 


3**11 




Sll-ll 








IMi 




8"lf 

a 6* 




S"|l 




h 


! 


Over- 
story 
Cover 


— . 
+ 


High 
Level 
Light 




111 




B»S 


: 


B-8 


t 


B-S 




8 


+ 
+ 


2 


J 1 


B* 


+ 


I 


I 

i « 
If f 

E 1 2 

la 1 



645 



o 

2 E 

E vi° 
ox:©' 

-= | j= 

C i- c. 

DUO 

■« i T3 

*- > >— 
<D -> U 

a v a 

£ © £ 



— (N m 

y u u 
q q a 



o § 

1 1 fc 

1 I i 

C — i p 

« 9 S 



1 

u 

Q 

u 

1/5 

u 

JC 
B. 

u 

Q 



O 



u 
o 

C 

o 



i 



o — < </-> o 

O O O — i 

© © o © 
VI VI VI VI 

0-. Oh Oh Oh 



> : 

a . • ■ *■" 

s + + + + 



u 
T3 
O 



u 

4- 



Ou 
<D 
U 
l— 

1> 
■*— 

I 

u 



u 
o 

c 

u 
VI 

x> 



u 

O 
E 

c 

u 
u 
JD 
v> 



c 
'3 



u 
O 
o 

c 
_o 

S, 

u 
i- 

o 



3 

E 
u 



A 

0. 



646 



I! 






•r £. ~ -% 

8-fIl 



60 


J 




i 




CO 








32 




TJ 


y—N 


O 


B 


1= 


o 


C 


m 


o 


© 


1/5 


VI 




D. 


u 






V 


D. 




•*-» 


o 


3 

s 


o 


ts 




c 


eo 


0> 


X! 


E 


CO 


c 


<L> 


o 

ha 


CO 


> 

c 


en 


1* 


D. 


4> 


TJ 


s 


C 


I 


o 


CO 


CO 


<!> 


-o 


ja 


c 


*-> 


J 


c 


c 






CO 


•o 


u. 




y 


cO 


■f 




U 


-«-» 


i 


c 
1> 


n 


CO 


E 

CO 


g 


D. 


a. 


C 


s 


CO 


CO 


9 


to 
4) 






.1 


8, 


i/j 


CO 




a> 


o\ 


l-c 

4> 


vn 


JS 


p 


£ 






-O 


CO 


CO 


(Tt 


H T3 



8-111 



llf-Jl 



8-JJ1 









i I - - i « 



= 1 

f- * 



III 



|J: 



111 

J - j 



8-8 



*" 



i i 



+ 

i 



O 5 






£ I 

s m 



i"s 



I 

I 



if 



T3 
<D 

3 
_C 

C 

o 
o 

I 

°^ 

SO 

1) 

I 



c 




00 


■ 

OS 






tk 

A 

a. 


d 


d 


d 


O 

8 

d 


§ 

d 


IS 


d 


p 


o 
(*> 
d 


d 


o 
d 


High 
Uvel 
Light 

X 

Tran- 
sect 




DC4 

X 
High 
Uvel 
Light 


i 

• 
■ 


DC2 

X 

High 
Level 
Light 




sil-ii 


i 


2-11! 


> 


DCJ 

Low 
Levd 

Light 








Over- 
story 

Tran- 
sect 




ii 




t 


; 


I 




ill 

« B 






i 


J 




High 
Uvel 
Light 


i 


Low 
Level 
Light 


t 


8-8 


i i 


8-8 


■ 
i 


g-8 




8 




+ 


i 


^ 


+ 


8 


i 




$ 


i 


1 


H" 




I 

■ 


■ 

1 1 


1 
1 


E 

r! 

2 1 


II 


j 
If 

8 S 

hi5 



647 



E £ 
-•no 



<n 


o 


o 




o 
o 


VI 


VI 


C/5 

o 


VI 


s, 






-c 


a> 


<u 


£ 




T3 


"O 








T3 


« 


La 


<s 


■*-» 




O 

c 




$ 




1 

V 


V 


V 


S3 

o 


E 


E 


l£ 


£ 


<r> 


>n 


5 


o 


i — i 




o 


© 


o 


'— 



U U U 
Q Q Q 



P 



1 

U 

Q 

v% 
<D 
c/j 
C/5 

U 



O. 
— 



C/3 



o o 



o — 



o o o o 
VI VI VI VI 

a. a. c_ a. 



> 
i 

8 

T3 

C 

CO 
CD 

a 
o 



c 
a 

Crt 

I 

O. 



<D 

! 

3 



4. 

*s I r. 

+ + + + 



rst 



-a 

_ u 
T3 cS 

£ -a 

S S 

£ -2 

"55 «fi 

D O 
— en 
O <D 

m s> 

1? c 
J> ca 

3 

3.2 

E +l 

-= •£ 
?« ft 

e s 

11 

>^ H 

3 2 

t+H > 

o <G 

c ' 

o 



648 



<D 

C 

2 
u 



"> 

T3 



-o 

c 

B 



g 

CO 

</) 
bo 

c 



T3 

u 8 

2 c 

? E 

(5 3 



Q O 




o o 




T3 C 




<U '~ 




t/i V) 




rt <D 




-O 3 








<n ea 




o> •> 




o " 




£, <o 




v> Os 




c <* 




S T 




'•8 (N 




S vo 




"S o 




60 — 




> c 




(m -r 




| 




22 "O 




B « 




•2 u 




■S3 S 




■2 « 




O l> 




o "3 




£ a. 




< § 




<z> 




O <_ 




— o 


d 


v© So 


o 


J. 1 


cS 


-2 « 


> 


l5 ~ 


u 

■a 







y"s 




on 
















allow, 
nstant 
+0.10m 




allow 
aceou 

ie 






.0 § 

5 -5 










JB O oo 




JC <-v u. 






•S 1 










00 U O 

© 




toS 
Herl 
Prai 






1 1 


i 
i 


,2 






5 a 5 


4 'f 


1 




Shallo 
Moder 
Variabi 


9 5 
o ■ 
o — 

o "". 
w o 










1 

lb 


t 

8 

B 


Woodward 
virginica 


I 

| 

1 










u .§ 






3 


^ 




Shallow, 
Variable 


4"? 






<=> 3 
<* 9 

Is 




to to Creek 
Moderately and Rive 
Deep Floodpla 
Herbaceous 


4) 

'C 

2 


Pima spp Saggiiaria Ilex myrtifo 
graminea 


Pehandra Bidens mitis Nyssa 
virginica ogeechee 


Eleocharis Lycopodium 
baldwinii/ spp. 
vivipara 






Moderate 

Depth, 

Constant 


(0.12-0.15+ 
0.05-0. 10m) 


Triadenum 
virginicum 


D 

IS 

E 

o 

1 


Depth, 
Moderate 
Variability 


(0.15-0.16+ 
0.14-0.20m) 






erate 

ath, 

able 








Mod 
De 

Van 






k to Deep 
r Herbaceous 
Prairie 






to 
S § 

5 g 


1-8 
« 8 


B 

s 

-J <3 






11 

O 


(0.23-0.30+ 
0.06-0. 10m) 


Andropogon 
virginiana 




u ^> 


*«? 






£13 






a a> «" 

g > <u 






s 


1 




% 1 

■S -£ 

11 

IS 




De 

Mod 
Varia 


J- 9 




uSg 






j 1 

11 




tZ 




£S 


C 


a a u 






us Utric 
spp. 


Cleth 
ainifo 




■ 

es v^j 


? 






Deep, Vari 
(0.20-0.2 


0.21-0.28 


[■S3* 

c 
2 
H 


Aquatic 
Prairie oi 
Lake 






1 

} 


Cephalanth 
occidentalis 


Nyssa 
sylvatica v. 
biflora 


Fraxinus 
caroliniana 



649 



1 

t/3 





F 


1 


o 


B 


?. 





or, 


C 


o 



•s M & 
> « sg 

"3 "fi 1 

ail 



9 S 

>o i 
o — 



11 

1 '5 



o 1 ? 

— CO 

Is 



2*5 J 

'an 

2 D 



+19 

in t 

— o 

2" 2 

« © 



If 

ll 



lol| 



si 

.A 9 
So 



2 -5 4 
"S F 5 

2 Q > 



a. § 
u i 

a § 



m O 






5 

I 



3<£ 



<L> 

3 
_C 
■*-» 

c 
o 
o 



H 



s > 



+19 



1 



a 



CM 00 

6 9 

sS 



c 

II 






I 

L 



9 


^ 


„0 


<s 


jf 


& 1 


ll 


■2i 5 



£ .3 



li 



£ 



r a "a 



| 



■5 



8 s 



■s 
8 

a 



1 






s I 



If 



51 

3 1 



B 

s 



11 s 



1 

I 



i i 

g § 

Q -5 



Q. 

Q. 



in 

I 



|J 

11 



£1 



5 

5 1 



si 



Lyonia lucid a 



650 




'"und, 



at/on 




**>i/Vo 



'"unt/, 



■Hon 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation-*- % of time in shallow water). 

Figure 6-8. Distribution of sample points in observed and model-predicted relationships 
between species abundance (1993-1994) and inundation depth and duration (1962-1995). 



Acerrubrum 



651 




'"una, 



at/on 




100 Jp 



.* 



> <r 



'"UOd, 



ation 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* V. of time in shallow water). 



Figure 6-8 -continued. 



Carex walteriana 



652 




tounv, 



at/o n 



Note: % of time in deep water is represented at XY origin, and is calculated as 
100% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued. 



Cyrilla racemi flora 



653 




buna, 



at/on 




%Of1U. 20 

"° '"una, 



at/on 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8~continued 



654 



Decodon verticillatus 




/0 <*Ti m 



'"U/)tf, 



at/on 



Bidens mitis 




80 

60 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation'*- % of time in shallow water). 



Figure 6-8- continued. 



655 



Dulichium arundinaceum 




OfTJm 40 

"^e With 



Orontium aqua ti cum 



lnun «atio n 




100^- 






N ° ,n ^atio n 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation'*- % of time in shallow water). 



Figure 6-8 -continued 



Eleocharis baldwinii/vivipara 



656 




N °"">n dation 




% <*T ime 



Note: % of time in deep water is represented at XY origin, and is calculated as 
100% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



Eriocaulon lineare 



657 




'"one/, 



ation 




Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued. 



Gordon/a lasianthus 



658 




***** 



found, 



atio n 




**h No 



found, 



at/on 



Note: % of time In deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



Ilex cassine 



659 




/o <* Time 



found, 



ation 




,ni "Wat/on 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



/tea virginica 



660 




'nunc/, 



at'o n 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8- continued 



661 



Lacnanthes caroliniana 




100 x & 



& 



60 ^° 



/o ofTi^ 20 

' me ^ith Noln 



ation 




**t *o 



' nUf Klatio n 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued. 



Leucothoe racemosa 



662 




8(P 

ou 

No, ^nctation 




**h No 



found, 



ation 



Note: % of time in deep water is represented at XY origin, and is calculated as 
100% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



663 



Magnolia virginiana 




**h No 



'"una 



at/on 




100 ^ 
80 /P 



rf 



% ofT im 



'"unci, 



ation 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation-*- % of time in shallow water). 



Figure 6-8-continued 



664 


Nuphar luteum 


*°*\~ — -— — ^__ 




u 




e 




8 60 1 




1 1 




* 


1 1 






"O 






1 4oi Jl 


1 1 


■ 




I ^L ~~P — ' 






i ^C^^- 




s\ ~~~—~?-x-^Ml^Bt «< 


20 \ /^Z2l3^& 


i L 








80^^ — y- *» »W~ 




i If 


ifr 4 ° ^ 


60^- _/ "'T- 

40^ — — -^ 




<F 20 ^ 


/0 ° fT ^e^ 


*•*-**„ ° ° * 


70 T^^ 




. 60 -i 


• _ J m 


o 1 


■ 


i ' 


11 T 7 ( 1 II 

1 1 1 f f if 


> 




_ 




o 50 i 


t. J 




f]_f I I ■£/ U ml 


* 40 1 ] 1 1 


1 


< 


i 


1 T 4 n 
jr i 


I 

J f J' 


IL IT 

Iff:;. 




<» • 1 












If ffP 






*i IT 


li 










1 Ui 






1H It t 


> f ^^1 


^Z. 






T 1 Jit 1 

i in 1 1 i 


|M Hlji 






■ I LW k 


< ^ 












[MfHO~§^H 




fc 201 I! 1 Jl 


, £j 












WW IHr 10C V vC< 


10 1 IMl 






'Mr 






Uflfi 1 ' ">^ 


lL ' 


IP; 


;HHfcjld Kt^ so ^ u 


"^So-^C^^F 


III II [HifHlyHff^ c^ 


60 ■ — — ~^r ■■ ■ i i|K — »-!■ ■£ ©• 
'"■"Wal/on 


Note: % of time in deep water is represented at XY origin, and is calculated as 


100% - (% of time with no inundation* % of time in shallow water). 


Fieure 6-8-continued 














i 



665 



Nymphaea odorata 




'"und, 



at/on 



Note: % of time in deep water is represented atXY origin, and is calculated as 
1 00% - (% of time with no inundation'*- % of time in shallow water). 



Figure 6-8- continued 



Nyssa sylvatica v. biflora 



666 




**h No 



found, 



athn 




100 up 



% of T ini 



found, 



at/on 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued. 



667 



Pan/cum hemitomon 



100 jp 

80 f 
60 ^° 



6 




% <*T7o, 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



Peltandra virginica 



668 




**hNo 



found, 



athn 




% <XT ilne 



W *h No 



lr >und, 



a«o n 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8- continued. 



Pier us phillyreifolia 



669 




% <"Tin, e 



/nu ncfat, ( 



Note: % of time in deep water is represented at XY origin, and is calculated as 
100% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8- continued 



670 



Rhynchospora inundata 




**h No 



'"unci, 



athn 




°v 



4 



% ofTi m( 



*ith No 



,r »ond, 



at/on 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8 -continued 



Smilax lauri folia 



671 




n*h No 



,r >und, 



at/o n 




Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



Smilax walteri 



672 




1e "«th No 



Inu 



"dat/, 



on 




**h No 



'"unci, 



a«on 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation'*' % of time in shallow water). 









Figure 6-8-continued 



Sphagnum species 



673 




'"una, 



at'on 




% °fT, m 



"""Wat,, 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continned 



Taxodium ascendens 



674 




*ith No 



,n unci, 



athn 




%0,Tln »»<toNo ln 2 ° 
*° 'nuna, 



at/o n 



Note: % of time in d««p water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation-*- % of time in shallow water). 



Figure 6-8-cnntinneH 



Utricularia species 



675 







,n und, 



at/on 




*'«iA/o 



'^uncy, 



ation 



Note: % of time in deep water is represented at XY origin, and is calculated as 
100% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8- continued 



Woodwardia virginica 



676 




20 ^ 



'nund, 



at/on 




0/0 °n ini( 



WthNo 



,n und, 



ation 



Note: % of time in deep water is represented atXY origin, and is calculated as 
1 00% - (% of time with no inundation'*' % of time in shallow water). 



Figure 6-8-continued. 



677 



Xyris species 




**hNo 



'"unci, 



ati Qn 



Note: % of time in deep water is represented at XY origin, and is calculated as 
1 00% - (% of time with no inundation* % of time in shallow water). 



Figure 6-8-continued 



678 
abundance where water depths were shallow 80-100% of the time, with no inundation 

< 20% of the time, and therefore by difference, in deep water < 20 % of the time. 
Species abundance at a site under this hydrologic condition is indicated by the height of 
the point and drop-line above the triangle base. This relationship is labeled A in the top 
plot of Lyonia lucida in Figure 6-8. At B in this plot, no inundation occurs 80-100% of 
the time, shallow water depths occur < 20% of the time, deep water depths occur < 20% 
of the time, and abundance of hurrahbush is low (< 10 stems/m 2 ). Points in the region 
marked C are shallow or without inundation, but never deeply inundated, and stem 
density gradually increases with increasing duration of shallow flooding. Points in the 
region marked D are always deeply flooded, and stems are absent or in low densities. 
Points in the region marked E are usually without inundation or in deep water, and in 
shallow water < 20% of the time; stem densities are < 10 stems/m 2 . The bottom plot 
illustrates the model-predicted Lyonia lucida abundance. Comparison of the observed 
(top plot) and predicted (bottom plot) abundances illustrates the model fit to the data 
(Rm 2 )- Plot curvatures indicate water depth class interactions; greater curvature occurs 
with more significant interactions among inundation conditions, and also indicates water 
depth variability. Examination of the plots reveals groups of species with similar point 
distributions and abundance-hydrologic environment trends. These groups are listed in 
Table 6-11. 



679 



Table 6-11. Associations of vegetation species based similar 3-dimensional plots of 
modeled species occurrence and daily depth-inundation duration relationships at transect 
sample sites during 1962-1995. 



Group 



2 
3 
4 

5 

6 

7 
8 
9 

10 

11 



Species 



Significant Model Parameters" ' 



chain fern, Walter's sedge, arum DC 1 , DC2, DC3, DC 1 xDC3, 

DC2xDC3, low level light, high 
level light, overstory cover 

beakrush, hat pins, yellow-eyed DC2, DC2xDC3, low level light 
grass, water willow, tickseed 

swamp blackgum, dahoon holly, DC 1 , DC 1 xDC3 
sweet bay 

Sphagnum spp., pond cypress, 3- DC1, DC2, DC3, DClxDC2, 
square, golden club, spikerush, DClxDC3, DC2xDC3, low level 



spatterdock 

loblolly bay, red maple 

Virginia willow, red root, 
maidencane 

bladderworts 

water lily 



light 

DCl,DC2,DC3,DClxDC2, 
DClxDC3,DC2xDC3 

DCl,DC2,DC3,DClxDC2, 
DClxDC3,DC2xDC3 

low level light, overstory cover 

DC3, low level light 

hurrahbush, climbing fetterbush DC 1 , DC2, DC 1 xDC2, low level 

light 



Walter's greenbriar, bamboo 
greenbriar 

titi, fetterbush 



DC2,DClxDC3, 
DC2xDC3, overstory cover 

DC1, DC2, DC3, DC2xDC3, low 
level light, high level light 



' Significant model parameters are listed for all species in group, but all may not be 
significant for all species in group. See Table 6-8 for significant parameter for specific 
species. 

b Average daily water depths are represented in depth classes as: 

No inundation DC 1 water depth < m 



Table 6-1 1 --continued. 



680 



Shallow water 


DC2 


Shallow water 


DC3 


Shallow water 


DC4 


Deep water 


DC5 


Deep water 


DC6 


Deep water 


DC7 



m < water depth < 0.05 m 
0.05 m < water < 0.15 m 
0. 15 m< water depth < 0.30 m 
0.30 m < water depth < 0.60 m 
0.60 m < water depth < 1.00 m 
1.00 m> water depth 









681 
Species' Environments and Modeled Hvdrologic Chang es 

Hydrologic changes predicted with sill removal were compared with significant 
parameters of the species-environment multiple regression models to suggest vegetation 
changes possible with sill removal. Affects of the sill on the swamp hydrologic 
environment are predicted primarily for the central and western swamp (Figure 3-9). In 
most of this area hydroperiods will be shortened under deeper conditions, and high water 
depths will decrease with sill removal. The region close to the sill will also experience 
more fluctuations in water depths, with longer periods of exposed peat. These 
hydrologic changes will be sufficient to permit slight changes in vegetation composition, 
based on the species-environment model results (Table 6-12). However, other factors 
such as seed dispersal, changes in disturbance regimes, and competitive interactions 
among invading and established species might modify the outcome of these changes that 
are predicted based on the hydrologic environments. 

Discussion 

Species Associations an d the Hvdrologic Environment 

The vegetation of Okefenokee Swamp is a moving mosaic in a landscape that has 
resulted in part from periodic, unpredictable perturbations and competitive interactions 
among species (Hamilton 1984, 1982). Fire is the most frequent, extensive, and intensive 
disturbance occurring in the system, although small-scale storms and infrequent 



682 

Table 6-12. Predicted changes in biweekly water depth range (from depth classes) and 
variability by swamp region, predicted by the swamp hydrology model with sill removal 
(summarized from Figure 3-18). 



Range of Most Frequent 


Range of Most Frequent 


Affected Area of 


Average Biweekly Water 


Average Biweekly Water 


Swamp 


Depths with Sill in Place 


Depths with Sill Removal 




0.05 -0.60 m 


- 0.30 m 


Chase Prairie 


0.05 -1.00 m 


0.05 - 0.60 m 


Craven's 
Hammock 


0.15- 1.00 m 


0.05 - 0.60 m 


Floyd's Prairie, 

Sapling Prairie, 

Sapp Prairie 


0.15- 1.00 m 


- 0.30 m 


Cypress Creek 


0.15 -1.00 m 


0.15 -0.60 m 


Territory Prairie 


0.30 -1.00 m 


0.30 -0.60 m 


Billy's Lake 


0.30 -1.00 m 


0.15 -1.00 m 


Sill Gate Area 


0.30 -1.00 m 


0.15 -0.60 m 


SCFSP 


0.60 -1.00 m 


0.30 -1.00 m 


Suwannee River 
Narrows, Area 
Southwest of Sill 


0.60- 1.00+ m 


0.30 -1.00 m 


Sweetwater Creek 









683 
hurricanes creating local damage due to high winds also occur (see Chapter 4). 

Vegetation communities occurring in the swamp today have been present throughout the 
swamp's development and occur throughout the accumulated peat (Cohen et al. 1984, 
Cohen 1974, 1973a, 1973b, Rich 1984a, 1984b, 1979), indicating that the disturbances 
occurring throughout the swamp's history have created conditions within the tolerances 
of the current suite of predominant species. There is a limited suite of species in the 
swamp seed bank that are tolerant of the swamp hydrologic environments. Variation in 
the landscape through time is an expression of these differential tolerances and species' 
plasticities in response to environmental change. Species composition and relative water 
levels (above and below ground) change through succession in a predictable pattern with 
peat accumulation, and are characteristic to a degree for each successional stage (Deuver 
1982, 1979). Replacement of species results as conditions gradually become unsuitable 
for those present, but disturbances, primarily from fire, eventually occur that recreate the 
environments more suitable for early succession species tolerant of longer and deeper 
inundation. 

Although fire history results in spatial variability of vegetation communities, the 
swamp hydrologic environment is also spatially and temporally variable; this variability 
and subtle differences in hydroperiod and inundation depths are reflected in the species 
composition, associations, and distributions. Pesnell and Brown (1977) found 
hydroperiod to be the primary factor determining plant distributions in Lake Okeechobee, 
FL, and Richardson et al. (1995) found that long-term hydroperiod variables (mean 
monthly water depth, variance, and hydroperiod) accounted for more variation in and 



684 
higher correlation with vegetation cover in lake Okeechobee, FL, marshes than short- 
term hydroperiod variables. Deuver (1988, 1984) found that hydroperiod and fire are 
controlling factors of plant community composition and distribution in Corkscrew 
Swamp and Big Cypress Swamp, FL, and Gunderson (1994, 1992) found that the 
Everglades system is structured by hydrology, fire, and vegetation interacting at multiple 
temporal and spatial scales. . The Okefenokee Swamp contains areas of nearly constant 
shallow and deep water, as well as areas dramatically affected by seasonal variations in 
water depths and flooding durations due to precipitation and evaporation variability (see 
Chapters 2 and 3). Species associated with each of these environments are related to 
different hydrologic conditions and histories, and can be modeled and diagramed to 
illustrate similar-species groups. The following discussion details these modeled 
relationships and compares them to species' environments in other southeastern 
wetlands. 

Hydrologic environments where most species occurred were not representative of 
conditions throughout the swamp where these species were absent. There were 
significant differences in inundation duration, inundation depth, and light availability 
conditions where species were present and absent across the landscape, and a gradient of 
differences in significant hydrologic parameters of all-sample models and species-present 
models. This indicates that many of the sampled species are specific in their hydrologic 
and light availability requirements; their distributions reflect a landscape that is 
hydrologically diverse, contributing to the heterogeneous landscape mosaic. 



685 
The long hydroperiod, deep water environments found in the aquatic prairies and 
the sill impoundment areas include associations of species that are differentiated by 
slightly different hydroperiods, water depths, and light availability variables (Tables 6-10 
and 6-11). Fragrant water lilies and golden club are found primarily in long hydroperiod, 
moderately variable, deep water environments, where daily water depths averaged > 0.30 
m during 40% of the time (1962-1995), and completely exposed conditions occurring 
occasionally (10% of the time). Although these species are also found at shallow depths 
and in partially shaded conditions, greatest coverage was in open water > 0.30 m deep, 
where ground level light is abundant. These conditions indicate inundation for longer 
periods than those found by Duever (1982), who estimated that fragrant water lily 
environments along the Okefenokee Swamp edge were without standing water for 44% 
of the 1941-1976 sampling interval. Richardson et al. (1995) found that fragrant water 
lily occurred in Lake Okeechobee, FL, in areas inundated for 93% of the study period 
(1970-1992), and David (1996) estimated a 96% inundation frequency with 61.5 cm 
average water depth for this species in Water Conservation Area 3 A of the Florida 
Everglades. Wood and Tanner (1990) did not find fragrant water lily in association with 
a tall growth form of sawgrass (Cladiumjamaicense) that occurred in areas with long 
hydroperiods and deep water, but found it in wet prairies characterized by deepest water 
depths and possibly shorter hydroperiods. 

Three species of bladderwort (purple bladderwort, Utricular ia purperea, floating 
bladderwort, U. inflata, rush bladderwort, U.juncea) were found in the understory plots 
where water depths were > 0.30 m for 42% of the time, with most frequent conditions 



686 
0.30-0.60 m deep; water depths were < m only 9% of the time. Purple and floating 
bladderworts were in greatest abundance in deep water conditions where ground level 
light was abundant and overstory cover was sparse; rush bladderworts occurred in 
saturated but usually not deeply-flooded sites, also where light was abundant (Bosserman 
1983a). These differences probably contributed to the non-significance of depth-class in 
modeled descriptions of sites frequented by bladderworts, and resulted in a 3- 
dimensional plot that differed slightly from its frequent associate, fragrant water lily 
(Figure 6-8). David (1996) found similar variability in inundation frequencies (61.4- 
96.4% of 7 years) and average water depth (37 cm) where bladderworts occurred in the 
Everglades, and Richardson et al. (1995) estimated that sites with bladderworts were 
inundated during 97% of the 23-year study interval in Lake Okeechobee, FL. Although 
Wood and Tanner (1990) indicate that bladderworts occur in wet prairie environments of 
the Everglades, they did not quantify the extent or duration of flooding where this species 
occurred. None of these studies included species identification of encountered 
bladderworts. 

Two other associations of species frequently occurred in long hydroperiod 
environments of deep water prairies, where water depths were slightly more shallow 
(0. 15-0.60 m) and peat surfaces were more frequently exposed (13-25 % of the time). 
Chain fern, white arum, and Walter's sedge occurred in greatest abundance where water 
depths were 0. 15-0.30 m or 0.30-0.60 m at least 50% of the time, and the peat surface 
was exposed approximately 20% of the time. These species occurred in sites with 
slightly shorter hydroperiods and shallower water depths than fragrant water lilies, 



687 
golden club, and bladderworts, and they were negatively associated with high light 
intensities. This contrasts with the results of Lucansky (1981), who found chain fern most 
abundant in sunny locations in North Florida. Wood and Tanner (1990) suggested that 
the occurrence of ferns and vines at the base of tall sawgrass in the Everglades area was 
in part due to the shallow water levels condition relative to the general wet prairie 
environment, and also due to shade provided by the sawgrass tussock. 

The 3-dimensional plots of spatterdock, 3-square, spikerush, and Sphagnum spp. 
were similar to that of chain fern, white arum, and Walter's sedge; although the average 
conditions where spatterdock, 3-square, and Sphagnum spp. occurred were variable deep 
water, most-frequent occurrences were in water 0.05-0.30 m deep (spatterdock slightly 
deeper for longer period), with peat surface exposure approximately 20% of the time. 
The higher average water depth is most likely due to occurrence of these species in the 
sill impoundment area. Elsewhere water depths were considerably more shallow, which 
explains the position of spikerush in the moderate depth-moderate variability group. 
During 20-30% of the time, abundances of these species were associated with water 
depths < 0.05 m. This finding suggests that although these species are found in greatest 
abundance when water depths are generally 0. 15-0.60 m, they can also tolerate variability 
in water depths, persisting during short-term drawdown when the peat surface is exposed. 
Spatterdock, 3-square, and spikerush were abundant in the seed bank of areas where they 
were also found in the standing vegetation (Chapter 7), in addition to spreading 
vegetatively or sprouting from rhizomes or tubers (Masters 1974), which may expl* 
their return following inundation after exposure in variable environments. 



lam 



688 
Richardson et al. (1995), Gunderson (1994), and Wood and Tanner (1990) 
characterized the spikerush (E. cellulosa and E. inter sticta) environment of the 
Everglades as "wet prairie", with average water depths of 77.5-93.8 cm and inundation 
occurring during 96% of the 23 year study period in Lake Okeechobee, FL (Richardson et 
al. 1995); Lowe (1986) measured an inundation frequency of 87-91% where Eleocharis 
spp. occurred in an East-central Florida marsh. Although these hydroperiod estimates are 
similar for those estimated in Okefenokee swamp for spikerush, the inundation depths 
are much greater. Light availability significantly influenced abundance of spikerush, and 
did not significantly affect abundance of spatterdock, Sphagnum spp., or 3-square. Shade 
inhibited growth and seed production of E. obtusa in experimental studies (Maillette and 
Keddy 1989); spikerush, which has a caespitose growth form similar to that of E. obtusa, 
may grow radially in response to low and patchy light levels, perhaps in an attempt to 
distribute the available light among all shoots (Maillette and Keddy 1989) while avoiding 
competition for space with vertically spreading, shade-intolerant species. 

Sphagnum spp. occurred on sites exposed 20.4 ±21.9% of the time. This genera 
has numerous adaptations for tolerating periodic drought (Andrus 1986), and varies 
growth form and rate with availability of moisture, light, and competition (Li and Glime 
1990). Sphagnum spp. are capable of regenerating under exposed or inundated 
conditions and may occur in mats that have survived for many decades (Clymo and 
Duckett 1986), indicating that they are most likely capable of tolerating short-term, 
seasonal variability in the hydrologic environment that might leave the peat surface 
exposed. 



689 
Yellow-eyed grass, beakrush, hat pins, water willow, and tickseed were common 
in herbaceous prairies and aquatic prairie fringe, and occurred in relatively constant to 
moderately variable conditions where inundation depths were usually 0.05-0.30 m, and 
frequently 0.05-0. 15 m. Peat surface exposure occurred < 1 5% of the time where these 
species were found. Other commonly co-occurring species are listed in Table 6-10. 
David (1996) found beakrush in the Everglades where water depths averaged 14 cm and 
inundation occurred during 53% of the 7-year sample period, and broomsedge (also 
occurring in Okefenokee Swamp herbaceous prairies), where water depths averaged 13 
cm and inundation frequency ranged 0-100%. Broomsedge and beardgrass {Erianthus 
giganteus) were present historically in wet prairie areas sampled by Wood and Tanner 
(1990), but absent from their sampling. They attributed this change to prolonged 
flooding; however, depth of flooding may have been an equally significant limitation. 
Richardson et al. (1995) found yellow-eyed grass where hydroperiods were shorter and 
inundation less than areas occupied by water lily-bladderwort communities. No 
abundant species in their study occurred where hydroperiods were < 75% of the 23-year 
study period. Beakrush occurred in the marsh of a Florida lake margin, where it was 
inundated during 94% of 1971-1981 (Lowe 1990). Gerritsen and Greening (1989) found 
beakrush abundant in an Okefenokee Swamp aquatic prairie seed bank, but sparsely 
occurring in the standing vegetation. Its greatest abundance occurred during droughts, 
when beakrush seeds in the seed bank responded to exposure by germinating, and a new 
cohort of seeds was produced (Gerritsen and Greening 1989). During non-drought years 
it occurred sparsely on margins of islands. Rather than continuously compete for 



690 
resources with shallow water, long-hydroperiod species, beakrush may take advantage of 
occasional extreme conditions (drought) and rely primarily on rapidly producing 
propagules that may still be viable after 400+ years of submergence (Gerritsen and 
Greening 1989, Conti and Gunther 1984) in areas with longer hydroperiods and deeper 
inundation. Light level also affected abundance of beakrush; where abundant, ground 
level light correlated with greater amounts of beakrush. Other modeled herbaceous 
prairie species (yellow-eyed grass, hat pins, water willow, tickseed) were not 
significantly correlated with low level light availability. 

Peat exposure occurred less often in the herbaceous prairies than in aquatic 
prairies, although herbaceous prairie water depths were on average shallower than those 
in the aquatic prairie environments (Table 6-4). This suggests some mechanism of water 
retention, by reduced run-off, minimal percolation, or low evapotranspiration rates. The 
overall topographic gradient across the swamp is from high elevations the Northeast to 
low elevations the Southwest. Along this trend are regions with perched surface water 
and minimal lateral water movement, creating a terracing effect in the water surface 
across the landscape (See Chapters 2 and 3). Durdin Prairie is in one of these terraces, 
which may partially explain the long hydroperiods in this area. Many areas of the swamp 
with herbaceous prairie vegetation also have flocculent peat and lack the firm peat 
bottom surface found in most aquatic prairies in the swamp. Flocculent peat occurs 
throughout Durdin Prairie, parts of Sapling and Floyd's Prairie in the vicinity of the 
Suwannee River floodplain, and some perimeter areas of Mizell and Chesser Prairies, 
and is likely in other areas not sampled. More frequent drawdown in the aquatic prairie 



691 
environment leads to compaction and oxidation of the peat surface, not experienced by 
the more continuously inundated surface in herbaceous prairie environments (Damman 
and French 1987). Along a transect in South Chesser Prairie, an up- welling of water 
(possibly a spring) was located within a few meters of an area of flocculent peat; this up- 
welling may maintain inundation in this area, even during periods of drought, creating 
this flocculent peat. 

Wetland plants have mechanisms to acclimate to stresses of a flooded 
environment, such as cessation of gaseous exchange imposed by inundation. Formation 
of adventitious roots, aerenchyma tissue, hypertrophy of stem lenticels, secondary root 
formation, and formation of knees or pneumatophores are be structural changes to 
increase exchange of oxygen and waste products occurring in response to flood stress 
(Kozlowski 1984a, 1984b). Many plant species occurring in Okefenokee Swamp have 
these features. Many of the abundant herbaceous prairie species are monocots with 
tough leaves and parallel venation; this leaf structure decreases evaporative loss (Cherrett 
1968). This feature may consequently increase the duration of flooded conditions in 
herbaceous prairies by retarding regional water loss due to evapotranspiration. Parallel 
venation also creates a leaf with high fiber content, which slows decomposition rates 
(Damman and French 1987). Dead vegetation accumulates in the absence of fire, and 
peat accumulates, eventually decreasing inundation of the peat surface. Therefore, 
species composition and abundance of peat in the water column may affect formation of 
floating islands and mats of vegetation, and subsequent succession from herbaceous 
prairie to wet forest. 



692 
Several species' groups occurred in the swamp where a variable hydrologic 
environment resulted in frequent peat surface exposure, although average water depths 
might be shallow or deep. Red root, maidencane, and Virginia willow have similarly 
shaped modeled surfaces indicating common relationships among variables and 
occurrences, with maidencane occurring at deeper and less frequently exposed sites. Red 
root and Virginia willow were found where peat surface was exposed nearly a third of 
the time, and the remainder of time most often inundated to 0. 15-0.30 m. Virginia 
willow stems were usually located on hummocks or bases of trees, stumps, or decaying 
logs, giving them slight elevation above the surrounding inundated peat surface. 
Although maidencane had a similar model-surface shape, no flooding durations in 
particular depth classes were significantly related to maidencane cover. However, 
interactions of occurrences in shallow water, deep water, and no inundation were 
significant, indicating a variable hydrologic environment. David (1996) found 
maidencane in areas less frequently inundated (61%) but at similar water depths in the 
Everglades, whereas Wood and Tanner (1990) found maidencane in wet prairie 
environments with deep water and long hydroperiods, similar to sloughs described by 
Gunderson (1994). Lowe (1986) considered 87% inundation frequency optimum for 
maidencane coverage in a North-Florida lake margin, and Richardson at al. (1995) found 
maidencane at sites with a 23-year inundation frequency of 98%. This variability in 
water depth tolerance illustrates the plasticity of this species, which probably enables it 
to rapidly colonize recently exposed peat and then persist as standing vegetation and in 
the seed bank in the understory of developing islands (Cypert 1972), as well as in floating 



693 
mats in deeply inundated areas. Light levels did not significantly affect abundance of 
these species under fluctuating hydrologic conditions in Okefenokee Swamp. 

Most tree and shrub species were found where exposure occurred about 25% of 
the time, and inundation depths were frequently 0.05-0.30 m. Abundant woody species 
were grouped into 5 associations based on shapes of 3-dimensional plots (indicating 
occurrence in exposed, shallow, and deep water conditions), water depth variability, and 
inundation depth. The associations are distributed along a gradient of water depth and 
variability. The moderately variable condition on the gradient is represented by pond 
cypress and red maple. These species had similar model-surface shapes, with red maple 
and pond cypress in slightly shorter hydroperiods and deeper water than loblolly bay, 
which has a similar modeled shape, but occurs in more constant conditions. Surface 
curvature in the 3-dimensional plots indicated significant interactions between frequency 
of exposed conditions and shallow water depths for red maple and pond cypress; 
abundances of red maple and loblolly bay were also higher with less time spent in deep 
water than pond cypress. Red maple is not normally found where deep flooding occurs 
during the growing season (Penfound 1952). Patrick et al. (1980) estimated that red 
maple generally occurs where soils are temporarily saturated or inundated for short 
durations (10-50%) of a year. Red maple in the Okefenokee Swamp averaged 40.0% + 
17.9% inundation duration to 0.08 m + 0. 19 m average water depth. Monk (1966) 
measured similar relationships among red maple, loblolly bay, and pond cypress in 
North-central Florida hardwood swamps, and Harms et al. (1980) recorded significant, 
increasing mortality in red maple where water depths were greater than 25 cm. After 6 



694 
years of flooding to an average daily water depth of 66 cm, red maple mortality was 22- 

100% in the impounded Ocklawaha River, Florida (Harms et al. 1980). 

The ability to produce adventitious roots when prolonged flooding occurs is 
related to flooding tolerance (Hook and Brown 1973). Although red maple and pond 
cypress can produce these roots, they did not do so everywhere in Okefenokee Swamp; 
adventitious roots were present on pond cypress only in the immediate vicinity of the sill 
and were not found on red maple in any of the sampled area of the swamp. Soils must be 
nearly continuously inundated to necessitate development of adventitious roots in most 
species with this capability; infrequent or annual periodic flooding does not usually result 
in their production (Harms 1973, Hook et al. 1972, 1971, 1970). This suggests that the 
level of flooding experienced by these species in areas outside of the impounded area 
(i.e., radiating north from Billy's Lake) in the Okefenokee Swamp was not prolonged 
enough to initiate production of these structures used in supplemental aeration. 

The model surfaces that describe the relationships among the hydrologic 
variables and occurrences of swamp blackgum, sweet bay, and dahoon holly are similar 
to those of red maple, pond cypress, and loblolly bay. Significant model parameters 
suggest that red maple, pond cypress, and loblolly bay densities increase with slightly 
longer duration of shallow flooding, whereas swamp blackgum, sweet bay, and dahoon 
holly abundances increase with increasing length of surface exposure. Peat was exposed 
25-30 % of the time where blackgum, sweet bay, and dahoon holly occurred. Swamp 
blackgum occurred where water depths fluctuated between no inundation and shallow 
flooding, with most frequent water depths < 0.60 m; although it can withstand deeper 



695 
inundation, blackgum productivity and recruitment are reduced by prolonged flooding 
(Harms et al. 1980, Patrick et al. 1980, Gill 1970, Monk 1968). Over a 6-year period, 
swamp blackgum mortality on the impounded Ocklawaha River was 20-55 % when trees 
were inundated to 125 cm (Harms at al. 1980), and remaining trees in 82-107 cm of 
flooding were in poor condition but appeared to have survived the long-term inundation. 
Red maple mortality approached 80% under those conditions (Harms et al. 1980). Sweet 
bay and dahoon holly abundances were significantly greater on Okefenokee Swamp 
sampled transects when water depths average < m. 

Although the modeled surfaces illustrate similar trends among woody species' 
occurrences (Tables 6-8, 6-9, 6-10, and 6-11), there are subtle differences that distinguish 
where these species might occur. Fetterbush, hurrahbush, bamboo greenbriar, Walter's 
greenbriar, titi, and climbing fetterbush were found primarily at sites with average water 
depths 0.06-0.1 1 m, moderate exposure, and most frequently between no inundation and 
water depths of 0.05-0.30 m. Hurrahbush and fetterbush occurred at moderately variable 
(SD = 0. 1 lm) shallower sites (most frequent water depths for hurrahbush 0. 15-0.30 m, 
and for fetterbush 0.05-0. 15 m). Fetterbush occurred more often with abundant ground 
level light, whereas sites with hurrahbush were generally more densely vegetated and had 
minimal ground level light. When considering only shallow water depths, duration of 
inundation was a significant parameter in predicting hurrahbush abundance, but 
fetterbush abundance was more significantly limited by abundant overstory cover and 
scarce ground level light. This finding agrees with Hamilton (1984, 1982), Deuver and 



696 
Riopelle (1983a, 1983b), and Cypert (1961), who recognized fetterbush as an early 
woody colonizer of exposed peat, and hurrahbush as a later succession, midstory species. 

Titi and Walter's greenbriar are found in wetter environments than fetterbush and 
hurrahbush. Walter's greenbriar is found primarily in water depths of 0. 15-0.30 m, while 
titi also occurs at depths > 0.30 m. Locations where Walter's greenbriar occur have a 
higher exposure frequency than those with titi. Abundances of neither species were 
significantly correlated with light availability. Walter's greenbriar is found in the 
understory in canopy gaps as well as in the canopy of low shrubs. Titi frequently forms 
the canopy in early stages of peat-based island formation, and is gradually replaced by 
hurrahbush and fetterbush as the vegetation ages (Glasser 1986, 1985, Best et al. 1984, 
Hamilton 1984, 1982, Deuver and Riopelle 1983a, 1983b, Deuver 1979), although it may 
persist in patches of sparse overstory. 

Bamboo greenbriar is found in areas of lower inundation depths but less frequent 
exposure than Walter's greenbriar. Greatest densities occur where water depths are 0.05- 
0. 15 m and overstory cover is abundant. Bamboo greenbriar occurs in much greater 
density than Walter's greenbriar in the swamp, and frequently grows into the tree and 
shrub canopy, creating an impenetrable blanket of vines across crowns of the woody 
species that give it support. It generally appears earlier in successional development than 
Walter's greenbriar (Cypert 1961). 

Climbing fetterbush is a unique species in the swamp; it usually does not root in 
the peat directly but grows in the crevices of pond cypress bark, and therefore occurs 
with a range of exposures and water depths. Its modeled surface most closely resembles 



697 
bamboo and Walter's greenbriar, which also use shrub and tree growth for physical 
support. Unlike the greenbriars, however, climbing fetterbush does not occur in 
abundance in the canopy; it is in greater densities where ground level light is abundant, 
which frequently is under a dense, high canopy. Stem densities are highest at sites that 
are frequently inundated > 0.30 m, which does not correspond to highest densities of 
pond cypress (0-0.30 m water depth). Overstory cover may exclude climbing fetterbush 
where pond cypress densities are high. This species is not usually found in early 
successional stages. 
Vegetation Changes Due to Sill Impoundment Effects 

The hydrology model predictions of areas in the swamp that are currently 
experiencing increased water depths and inundation durations were discussed in Chapter 
3 and are delineated on Figure 3-9. The increasing hydroperiods and water depths due to 
the sill impoundment have created an additional environment in the swamp close to the 
sill structure that was not previously present in that area, and species associated with this 
environment are currently unique to that part of the swamp. Prior to sill construction, 
the region directly north and east of the sill was a seasonally flooded pond cypress- 
swamp blackgum-pine forest with myrtle-leaved holly (Ilex myrtifolia), loblolly and 
sweet bay, and red maple scattered throughout. Marketable pines were removed from the 
area before flooding, and the area currently is vegetated with various-aged groups of 
pond cypress and swamp blackgum, with occasional bays, maple, and ash; pines occur 
only to the west of the sill where they were not logged during sill construction. Carolina 



698 
ash and ogeechee lime (Nyssa ogeechee) currently are found within the river floodplain 
from the sill northeast to the natural sill. Ogeechee lime is limited to the river banks and 
river floodplain above and below the sill structure, and also along the northwestern 
swamp beyond the area of the sill's influence in slow-flowing creeks. Carolina ash is 
scarce throughout the floodplain south of the natural sill, and infrequent elsewhere in the 
swamp; its distribution prior to sill construction is uncertain, although conditions were 
probably favorable for its occurrence in this area of the swamp. Penfound (1952) found 
Carolina ash in temporarily flooded flats and sloughs, frequently following fires, and 
probably transitional between swamp and mesic forest. Monk (1966) categorized 
Carolina ash with other mixed hardwood swamp species (sabal palm, Sabal palmetto; 
American elm, Ulmus americana; bald cypress, Taxodwm distichum), where flooding is 
seasonal and variable and peat accumulation is minimal. Monk (1966) found these 
species similarly distributed along a water depth-pH-cation gradient in North-Central 
Florida hardwood swamps; he suggested that the gradient represented a transition from 
hardwood and bayhead communities. Monk (1968) also found mixed hardwood swamps 
in vicinity of limestone outcroppings in Florida; Trowell (pers. comra.) believes there are 
similar outcroppings in the sill area within the swamp. Hamilton (1984, 1982) 
hypothesized that mixed swamp was a mature stage in the swamp in the absence of fire. 
In the remainder of the sill-affected area, extended flooding and deeper water depths 
favorable to aquatic prairie development have occurred, and germination of species 
requiring exposure has probably been reduced. However, flood-tolerant species that 



699 
could establish during drawdown in sufficient light (such as pond cypress, blackgum, and 
ogeechee lime) have also persisted. 

The region currently impounded by the sill would most likely encounter increased 
exposure duration and variability with complete sill removal, and experience an 
intermediate level of change with partial sill removal. In some of this affected area the 
degree of hydroperiod and depth changes would be sufficient to stimulate changes in 
vegetation composition. Competitive interactions among species, availability of 
propagules, and stage of successional development will modify this response. 

Changes in hydroperiods and water depths predicted in response to sill removal 
will not be uniform across the sill-affected area (Table 6-12). The region around the sill 
structure to Craven's Hammock and eastward to Billy's Lake, including the Suwannee 
River fioodplain will experience more variability in water levels with a decline in deep 
water depths (>0.30 m), and greater decline in depths and increased variability closer to 
the sill structure and creek and river channels (Table 6-12). Open canopy in the Craven's 
Hammock area will be more favorable for shallow prairie vegetation and shrub and tree 
reproduction. Although water levels will be more variable and therefore more favorable 
than current conditions for pond cypress regeneration, this species will probably be 
replaced in much of this area with loblolly and sweet bay and blackgum; although some 
cypress seed source exists in the area, bays compose much of the canopy and create 
shaded ground level conditions not tolerated by germinating cypress (Best et al. 1984, 
Demaree 1932). Cypress regeneration will be limited to large canopy gaps currently near 
seed source trees or within the area likely flooded by seasonal water, unless severe fires 



700 
occur which open the canopy and remove the surrounding bays and blackgum. However, 
severe fire could also reduce seedling survival (Cook and Ewel 1992). 

As predicted by the swamp hydrology model, water depths in the region bounded 
by Sapling Prairie southeast to Chase Prairie and possibly southwestern Territory Prairie, 
and southwest to Billy's Lake will decline < 0.30 m, with slightly more variability in the 
southwest, with sill removal (Table 6-12). These conditions will be more favorable for 
shallow, herbaceous prairie species, although aquatic prairie species will persist in areas 
with deeper water levels. Because much of this area is currently forested with bays, 
surface water flows are limited, and pond cypress is generally limited to prairie islands 
and isolated forest stands, there will probably not be a significant increase in pond 
cypress in this area, and areas forested in gums and bays will persist. In shallow open 
areas loblolly bay, which has wind-dispersed seeds, will probably be the most 
substantially increasing woody species. Currently loblolly bay is dispersed throughout 
the area, and seedlings are encroaching into eastern Floyd's Prairie from the Floyd's 
Island southwestern perimeter. Within this region, Floyd's Prairie will probably 
experience the greatest change, with increasing woody growth (primarily tit and loblolly 
bay)and less dramatic changes occurring in Sapling and Chase Prairies. Because 
Territory Prairie is terraced above Chase Prairie, change in hydroperiods that would lead 
to altered vegetation will probably be minimal; this region is currently affected by the sill 
impoundment primarily during periods of abundant precipitation, when water levels are 
high and constant. 



701 
The region bounded by Cypress Creek, Sapp Prairie, and Sweetwater Creek will 
also experience hydrologic changes with sill manipulation (Table 6-12). Cypress Creek 
will probably experience greater water level fluctuation and may experience extended 
exposure. Although Sapp Prairie and Sweetwater Creek may also have decreased high 
water depths, they will probably not experience the increase in exposure predicted for the 
Cypress Creek area. Sill removal will permit a greater volume of throughflow in the 
Suwannee River at the Cypress Creek-Suwannee River junction, which may slow 
drainage from Cypress Creek during high water events. However, during low 
precipitation periods, the river flow volume will decline more rapidly than with the sill in 
place, and flow from Cypress Creek will also rapidly decline. Vegetation changes in the 
area will probably be minimal, because most of the area is currently forested with shrub 
and wet pine communities and herbaceous prairie vegetation where openings exist. This 
area has experienced frequent fires since the sill's construction, maintaining these 
vegetation types which also occurred in the area before sill construction. Much of Sapp 
Prairie is slowly succeeding to shrub prairie; with sill removal, this trend will continue 
unless severe fires occur in the next decade. In the remainder of the swamp, vegetation 
changes that can be attributed to sill removal will be minimal; the absence of fire will be 
the primary function driving swamp succession. 



CHAPTER 7 

RESPONSE OF THE OKEFENOKEE SWAMP SEED BANK 

TO ALTERATIONS PN THE HYDROLOGIC ENVIRONMENT 



Introduction 

Accumulated seeds in sediments, or the seed bank, are a dormant reserve 
providing propagules for vegetation community establishment and maintenance, and 
recovery from extreme conditions and disturbances. The standing vegetation may be 
represented in this reserve, permitting perpetuation of the vegetation community as long 
as suitable conditions exist and seeds remain viable. Distinct patterns in the adult 
distributions may be mirrored in seedling and seed distributions; however, generalized 
tolerances of a broad gradient of conditions during recruitment may increase propagule 
and seedling survival and result in broad distributions of adult plants (Keddy and Ellis 
1985) so that zonation is not apparent. The established vegetation also may be persisting 
in conditions unsuitable for development of its propagules; changes in the site 
environment as the seedling matures may prohibit establishment of its own offspring at 
the site. The seed bank may also include seeds from species that have been removed 
from the site's standing vegetation through competitive interactions, succession, animal 
or human activities, or disturbances such as fire. Species that did not previously occur in 

702 



703 
the site's standing vegetation may also be present in the seed bank, transported to the site 
by water, wind, or animal movement. Survival and germination of these seeds depends 
on exposure to conditions that will break seed dormancy after deposition at a suitable site 
for seedling growth. When appropriate conditions for seed germination occur, 
competitive interactions and sensitivities to current and changing site conditions 
determine which seedlings will survive, reproduce, and contribute to the seed bank. 

Environmental conditions affect seed longevity and seedling recruitment from the 
seed bank (van der Valk 1981). Buried seeds in a wetland must survive anaerobic 
conditions in waterlogged sediments, and may have to tolerate inundated conditions as 
the seedling emerges. Those that can not germinate in flooded soils must wait for 
drawdown and soil exposure to occur. Changing nutrient dynamics, decomposition 
processes, and alterations in competitive interactions occur with increasingly aerobic 
conditions as water levels decrease and sediments are exposed. Seed bank compositions 
may serve as indicators of past wetland hydroperiods (Poiani and Johnson 1 989). Slight 
modifications in inundation depths or durations can alter the composition of standing 
vegetation (see Chapter 6) and subsequently seed bank composition; changes in 
successional patterns may then follow (van der Valk 1981). 

Seasonal weather patterns may create suitable conditions frequently (e.g., 
annually), so that perpetuation through long-term dormancy in the seed bank is 
unnecessary for some species. Those species persisting in the vegetative state can 
respond to short-term, favorable changes in environmental conditions through vegetative 
propagation or seed production. Alternating dry and wet conditions may also release a 



704 
suite of species unlike those persisting in continually dry or inundated conditions 
(Gerritsen and Greening 1989, Greening and Gerritsen 1987) and may affect survival of 
dormant seeds that need continuously inundated or exposed conditions for germination 
(Berrie and Drennan 1971). Severe or less predictable conditions such as drought lead 
some species to another strategy; these species persist as seeds that germinate only under 
infrequently occurring conditions, rapidly maturing and producing abundant seeds, and 
then residing again as seeds in the sediments, as competitive interactions with more 
persistent species increase (Grubb 1998). Elimination of the occasional disturbances such 
as severe drought and fire, or changes to the ambient hydrologic environment, may 
eventually displace species that perpetuate episodically in response to these 
unpredictable environmental fluctuations. Those that depend on general but predictable 
environmental conditions will also be displaced if environmental changes exceed their 
plasticity; a different community composition and landscape structure will eventually 
result. 

The Suwannee River sill was constructed to reduce frequency, extent, and 
intensity of wildfires by continuously flooding the swamp with the impounded Suwannee 
River, regardless of precipitation conditions The extended inundation duration and 
increased water depths resulting from the Suwannee River sill (see Chapter 3) have the 
potential to alter swamp vegetation community compositions and distributions locally in 
the hydrologically affected area (see Chapter 3 and Chapter 6). The sill may have 
affected the seed bank composition in this area by changing the composition of standing 
vegetation in response to continual flooding. In the sill-affected area opportunity for 



705 
recruitment of species requiring exposure for germination has decreased; potential for 
future expression of those species in the standing vegetation declines with continued sill 
operation, as seed viability in the inundated conditions decreases over time (Schneider 
and Sharitz 1988, 1986). 

Indirect, landscape-level changes may also be occurring in response to the sill 
impoundment; extended, deep flooding limits regeneration of species such as pond 
cypress {Taxodium ascendens) that require exposed surfaces for germination. Decreasing 
densities of pond cypress in the forest composition may affect response to wildfires, as 
fire-tolerating species such as pond cypress, which requires open canopy for seedling 
survival, are replaced by those with fire-suppressing characteristics (see Chapter 5). 
Pond cypress tolerates fire that may control or eliminate competing shade-tolerant and 
fire intolerant hardwood species in the surrounding forest. Although lack of wildfire in 
the impounded area can not be directly attributed to the sill impoundment affects, poor 
pond cypress regeneration due to extended hydroperiods and subsequent changes in 
species composition to less fire-tolerating or fire promoting species will create a positive 
feedback loop that eventually affects the area's fire regime. Reduction or elimination of 
pond cypress from the forest canopy is eventually possible if wildfires do not occur 
frequently and severely enough to remove its competitors (Best et al. 1984). 

The conditions of extended flooding have been exacerbated in the swamp by 
early 20th century logging activities (see Chapter 4), which removed or damaged much 
of the cypress and subsequently altered the regenerative potential of this community 
(Hamilton 1984, 1982). Decisions to alter the sill structure must consider the delayed 



706 
and long-term impacts of logging. Decreasing the inundation duration of the sill-affected 
area and permitting greater amplitude of seasonal fluctuations in water levels and 
flooding duration with sill removal will improve conditions for floodplain species that 
survived the early 20th century logging. However, species that have not persisted in the 
area's seed bank or standing vegetation, and that have not been brought into the area by 
wind or surface water movement, will be absent from the area until transported in as 
seeds or vegetative sprouts from other parts of the swamp and its perimeter. 

Dependence on the swamp seed bank to repopulate (to pre-sill composition) areas 
that might experience changes in the hydrologic environment with sill manipulation 
prompted the following questions, which are addressed in this chapter: 

1 ) What is the seed bank composition in the sill-affected area, and does it 
differ from that of the surrounding swamp? 

2) Do the seed bank contents differ from the site's standing vegetation 
composition? 

3) Is the seed bank composition representative of the site's hydrologic 
environment and recent history? 

4) What is the potential response of the seed bank to exposed and inundated 
conditions that might occur with sill modification? 



707 
Methods 






Seed Ban k Samplin g 

Emergence techniques (counts of seedlings germinating from seed bank samples 
housed in a greenhouse) were used to identify and quantify the swamp seed bank 
composition and response to hydrologic manipulation (e.g., Grillas et al. 1992, Poiani 
and Johnson 1988). During October (autumn samples) 1992 and 1993 and May (spring 
samples) 1993 and 1994 seed banks were sampled along transects in the 5 regions of the 
swamp where standing vegetation composition was assessed for species-hydrologic 
environment relationships (see Chapter 6). Half of the 16 transects in each area were 
randomly selected for seed bank assessment. These transects were also sampled for 
shrub and tree composition, as discussed in Chapter 6. Seed bank samples were 
collected within the structural zones identified for standing vegetation composition 
assessment. Collection of samples across the variety of structural zones existing in the 
sample areas provided an opportunity to examine differences in seed bank composition 
among succession^ stages (Leek 1989). Within each structural zone 6 cylindrical peat 
cores (approximately 20 cm deep x 20 cm diameter) were removed from sites proximal 
to the transect understory plots, within approximately 5 m of the transect and sampled 
plots (Figure 7-1). Sample depth was selected based on estimates of Gunther et al. 
(1984), that >90% of the viable seeds collected from wooded forest and open marsh 
areas of Okefenokee Swamp resided in the top 20 cm of peat. Several small samples 



708 



o 

5J 
<*) 

C 




111 

00 E ■ 

H trt 4) 

(75 w *"* 



60 

c 
o 

O 

o. 

<u 

"o. 

E 
eo 

c/i 

C 

o 



U 
eo 

> 
i- 
u 
x: 

o 

o 

13 



C/5 
4> 



c 
o 

4> 



o. 



£> 




T3 




§ 




m 




-— 




o 




<u 




E 




<L) 




J= 




U 


a. 


en 


E 


A 


ee 


11) 


5 


C 


C/J 


o 






.* 




o 


^"7* 


c 


r^ 


<s 


it) 


1) 


k. 


-j: 


3 
60 


o 


- 


,c 



709 
were anticipated to better represent structural zone seed banks than few, large samples 
(Thompson 1986). The 6 cores were combined in a large tub; live and undecomposed, 
dead materials were removed; and, the core material was thoroughly mixed to create a 
composite representative of the variability throughout the entire sampled zone. A 
subsample (approximately 4 1) was removed from the composite sample and stored in 
marked plastic bags for transport to the greenhouse near the Okefenokee Swamp 
National Wildlife Refuge, Camp Cornelia Visitor's Center, south of Folkston, GA. 
In the greenhouse (within 12 hours of sample collection) the 4 1 sample was 
halved and spread in a 2 cm thick layer in 2 plastic, potted-plant drainage pans 
(approximately 30 cm diameter x 7 cm deep) with perforated bottoms. Pairs of pans 
from each transect zone were randomly placed in the greenhouse in spillways (4, 
approximately 1 m x 12 m x 0.3 m) that were continuously flooded with swamp water 
pumped from the bottom of the boat basin canal near Camp Cornelia Visitor's Center. 
Water entered the spillways at the northwestern end, and flowed through the spillway to 
the south end, where it was gravity-drained back into the canal (Figure 7-2). One sample 
in each pair was placed on the spillway floor and the other was perched adjacent to it at 
the water surface on a brick. The brick held the sample approximately 4 cm above the 
spillway floor so that the sample surface remained moist by wicking water through the 
pan base without flooding the peat surface. The other member of the pair was inundated 
with approximately 2 cm of water that wicked through the pan base but did not overflow 
the pan edge. Thus the sample surfaces were kept isolated from the flowing water 
surface except through wicking through the pan bottom; this was intended to keep seeds 



710 




E> 

c 
o 



o 



a. 
E 

$ 



O 

3 

5 
u 



711 
that might be in the irrigation water from contaminating the sample surface. A pair of 
potting soil samples (2 1 in each) was similarly placed at the north end of each spillway to 
intercept irrigation water as it initially entered the spillways; these samples were 
intended to indicate the seed contents of incoming irrigation water. At the end of each 
sample interval, none of these trays contained seedlings of species found in the peat 
samples. Therefore, it was concluded that the seedlings emerging from the peat samples 
originated as seeds at the collection sites and were not carried in irrigation water. The 
continuously inundated or exposed treatments were selected based on Gerritsen and 
Greening's (1989) conclusion that germination of Okefenokee Swamp seed bank species 
was under either of these conditions. 

Although the greenhouse environment offered some protection for autumn 
samples from winter freezing temperatures, seasonal dynamics in air and water 
temperature and sunlight availability in the greenhouse generally reflected those in the 
surrounding swamp. Samples were not permitted to desiccate during periods of normally 
low water levels in the swamp. Surface peat at the collection site did not naturally 
desiccate during the study interval, although drawdown, peat exposure, and desiccation 
occasionally occur at some of the sample sites. Natural lighting was not supplemented 
during the germination periods. The duration of the germination period was determined 
by a general cessation of seedling emergence in September and February following 
spring and autumn sample periods, respectively. In September (following spring 
sampling) and February (following autumn sampling), all seedlings were removed from 
the sample trays, identified, and counted (Poiani and Johnson 1988). Seedlings that 



712 
could not be identified were retained in the sample trays until maturity made their 

identification more apparent. If seedlings matured to produce seeds during the sample 

period, the plant was removed from the sample tray at the peat surface to eliminate 

contamination with seed dehiscence in surrounding samples. 

Analysis of Seed Bank Emerg ence Data 

Descriptions of the transect herbaceous and woody vegetation and hydrologic 
environment were outlined in Chapter 6. Species composition in the understory plots 
was summarized across the zone, and understory plot hydrologic data were similarly 
averaged to describe the zone hydrologic environment where the sample originated. 
Counts of emerging seedlings were log-transformed (log 10 count +1) to normalize their 
distributions, and compared by seasons (spring, autumn) among sample areas (Chesser 
Prairie, Durdin Prairie, Sapling Prairie, Floyd's Prairie, Sill Area), structural zones 
(Table 6-1), treatments (submerged or exposed), and their interactions using a nested 
model ANOVA (Proc GLM, SAS Institute, Inc., Cary NC 27513). Parameters with 
significant effects were identified with mean comparisons using specified error mean 
squares. Overall comparisons between seasons were made with t-tests. Sample tray 
species diversity, species richness, and total seedling counts were similarly compared 
among areas, treatments, zone types, and seasons, as were species' groups based on 
general water depth and variability trends (see Chapter 6). 



713 



Results 



Species' Responses 

Forty-nine species (Table 7-1) germinated in the seed bank samples, representing 
a variety of vegetation community types in the swamp, although not all species in the 
standing vegetation were included in the seed pool (Table 7-2). Germination of woody 
species was sparse (Figure 7-3), which is not an uncommon feature of wetland seed 
banks (Leek 1989, Schneider and Sharitz 1986). Nine herbaceous species comprised 
93.4% of the germinated seeds; 71.2% of the emerging seedlings were yellow-eyed grass 
(Xyris spp.) (Table 7-1). Woody plants accounted for <0. 1% of the germinated seeds, 
and titi (Cyrilla racemiflord) was the most abundant of these species (Table 7-1). No 
species were found in the seed bank that were not also found somewhere in the swamp 
standing vegetation (although not necessarily along sample transects), and the standing 
vegetation composition at the collection site did not always correspond to the 
composition of germinated seeds from the site (Table 7-3). 

Total number of emerging seedlings did not differ among seasons, but did vary 
with area, zone, and treatment (Table 7-4). Diversity of the seed bank was also 
significantly affected by area, zone, and treatment, but seasonal differences were 
minimal (Table 7-4). Autumn species diversity was greatest in samples from tree, 
aquatic prairie, shrub-aquatic prairie, and tree-aquatic prairie, and in the spring was also 
high in samples from herbaceous prairie and aquatic-herbaceous prairie structural zones 



714 





Seasonal 
DifTerence 

In 
Variances 


e A 

II 0" 


\0 

d 


? 


? 


x> 


in 
r~ 

d 


i 

o 


d 




d 


g 

d 


? 


eo 

C 
G 

E 
| 

g 


■ 
T5 3 ■ 
See 

2 fc * II 

■ 5 -S 


A 

a. 


8 

00 

© 


o 

VI 

d 


o 

8 

d 


d 


oo 
d 


d 


oc 
d 


r- 

o 
o 

d 


d 


d 


d 


1 

co* 

c 
o 

CO 

Q 

CO 

co" 

2 


E j ffC 
ill* 


t^ 


d 




? 


CN 

d 


d 


d 


IN 


00 




<3N 


rite 

lis! 


r-- 


d 


d 


o 


d 


d 


d 






00 


vS 


«3 

ac 

c 
o 

E 

efl 
so 

C 

o 

•♦■-» 

3 
X) 

C 

CO 

T3 

'S 

X! 

c 

CO 
CO 

U 

D. 

E 

CO 


p 1 1 fee 

1 | =5 "1 -g 
to r- 


W) 


d 


d 


o 
d 


d 


d 


© 
d 






SO 




8 1 = ^^T- 
1 £ S 1 E 


00 


d 




i 


d 


d 


(N 






vd 


csi 


£ J -^^ 
i« <» O w 


d 


i 

d 


d 


d 


i 

d 


d 


d 


o 
d 




• 
>d 


d 


co &• ,8 o w 


d 


i 
o 

d 


* 


i 

o 
d 


■ 
o 

d 


* 

d 


o 
d 




*1 




a; 


M 

c 

c3 
X) 
T3 




Mill 


d 


i 

d 


• 


* 

o 
d 


i 

d 


d 


i 

o 
d 


1 




d 


* 

fN 


co 
U 

X! 


:= j .E ffO 

Hill 


* 


* 


1 


o 
d 


i 

d 


d 


* 
o 
d 


* 

si 


00 

d 




1 

o 
vS 


C 




d 


o 
d 


i 


i 

o 
d 


i 
d 


* 

d 


i 

o 
d 


1 

o 


* 

r-' 


d 
55 


r-; 


60 

CO 

3 




o 


Si 


d *^ 


'5 


?- 


2 TT 

O ~ 


^ ON 

o — 


d oo 


^8 


»N 00 




oo » 


i 

e 

X 

H 


5 

1 

CO 


1 5 


.<2 

i 

1 


.0 

31 


-£ .s 




s 9 


il 

Q 3! 


,<3 

El 

1 i 

Q 1 


|] 

5 1 


•S N 

1 1 5 
s J s 
•a a S 


5 £ 



715 



■tj 

3 
E 

o 



I 



af I I £ -i 
is a 



si 

£ J 



A 

a. 



Iff* 



rl 



1 J 



*ll* 



X> 



X> 



o 
o 



o 
d 



d 



d 



e 

c 



e 

c 



o 
d 



■a — 

Iff II 



111 !i 
fill!* 



K (« O w 



1 •£ ,£ ■&' 



be 

~ E 






£ £ I fc 
fjf (I 

fc ft- ^5 a * 



o 
d 



o 
d 



o 
c 



o 
d 



* 
o 

d 



* 
o 
d 



e 
a 



111 si 



HII» 



Sis eh IS 



J 



o 
d 



* 





d 



o 
d 



* 

o 
d 



■ 

o 
d 



c 

d 



q 

e 



d 



o 




o 
d 



o 
d 



to 

rn 
00 



* 
o 
d 



o 
d 



a 

d 



— O T 



£ B 
s ?* 



■S £ 

1« 60 



.2 S 
a -S 



I 

£ 



e 

I 

S 
3 



o m 



I 



* 

rr, 

d 



— ^ _ 

. «*> ^ o 
vo tn CD- 



'S .5 
S .5 



H 






■5 

.3 



716 



i 



© A 

II »• 



- s 



i e 



I S - - 

A 

0. 



is 

5 ^ 



E .5 I 




d 

d 



X> 



? 



? ? 



? 



c 

d 



e 

DC 



00 


00 


o 


8 


1 

c 


15 


o 


o 


o 



e 
d 



e 
d 



e ■- 

■c 5 



- E 



CO H 



111 || 

M c i o w 




o 
© 



q 
e 



00 

d 



o 
d 



* 



f •= -S £•£ 
<» ft- & a • 



01 



If * 

fa. a. 



« E 
1 £ 



IJ* 



B V 



at 



s ~ 



*jC 



1 1 f 1 1 



ii 



fc> *SS T3 



" E 



o 
d 



d 



o 
d 



o 
d 



o 
d 



i 



i 



o 
d 



&*$&* 



-a 
u 

3 
C 

+-» 

C 

o 
o 



1 

H 



S 1 1 J £ § g 



SB 



* 
00 



d 



* 

o 
d 



CI* 

d 



■ 






£ — 



? 



~8 



O ^T — — 






I 



Ii 







717 



1 



| J So a 

1 f i * 



9 ? 



I * 1 *" 

1 2 E i « 

5 .£ S ° 

*5 -E 






00 


oc 

00 


o 


c 



HI l 



o 
ci 



o 
d 



e 
d 




? 



? 



o n or 

ill ij 

X 



Mill 

111** 



O 

d 



e 
d 



o 
d 



o 
d 



c 
d 



e 
d 



lifts 



? 



? 



1 1 ttc 

ill 



o 
d 



o 
d 



o 

o 



e 



o 
b 



o 
d 






01 «, 

* 1 1 1 I 



ON 

d 



o 
d 



■ 

o 
d 



o 
d 



o 

o 



<D 

3 
C 

■*^ 

c 
o 
o 



.2 
H 



111 £ l 



* 
o 



o 
d 



o 

o 



o 
d 



* 
d 



e 
d 



i 

d 



o 

d 



■ *C — £ « 

^ — = "S ■ 



* 



* 

d 



* 
o 



2 A S *3! s? 

8*11* 



o 

d 



d 



I 

d 



o 
d 



B I f Si i * 



*$&* 



It 






?_ 



n o 



?. 






11 



.0 ^ 






«.*, 



c 

3 ! 



If 






4 




718 



-a 

u 
3 
5 

a 

o 



u 

1 



H„: 

Seasonal 

Difference 

in 

Variances 

= 0? 

P>t 


1 




(^ 


00 

i 


p 


* 
o 


? 


X) 


d 


d 


e 
d 


d 
















Seasonal 
Differenc 
in Means 

= 0? 

P>t 




r- 










■^T 


— 


<N 




r> 




o 


o 


O 


o 


o 


o 


tumn 
dling 
nsity 






<N 








33** 


$ 


- 


f*1 


r* 


■*r 


00 




o 


Os 


o 


f 


00 


ring 

dling 

nsity 


















r*i 








*ia* 


— 


o 




00 


*n 


"5 




o 


o 


00 


o 


TT 


F 


•o — 




























Submerg 

Treatme 

Seedlin 

Density 

(#/m') 














? 


d 


SO 
rr 


d 




On 


■O B H - 














Expose 

Treatme 

Seedlin 

Density 

(#/m 2 ) 


d 


o 
d 


in 


OC 

d 


vi 




Jill 






^ 








j *o " 


- 


o 




r« 


>C 






° 


c 


»N 


o 


f*} 


F 


flffc 

iiL '« 1 1 -S 














i 


1 


^3- 






fN 


r> III 

05 * i Q 9 


o 


o 




r~ 


00 


r- 




o 


o 


\C 


o 


rr-t 


00 


Floyd's 
Prairie 
Seedling 
Density 

(#/m') 


o 


1 

o 


O 


NO 


(S 


\0 




d 


o 




o 


^T 




1 -c f £C 






* 








ill II 


* 


* 

IN 




r» 


00 






o 


o 


~* 


o 


■*T 


ts 


Chesser 
Prairie 
Seedling 
Density 

(#/m ! ) 














1 


1 

o 


1* 


o 


oo 


CO 




o 


o 


WD 


~ 


t 


OS 


2 — — C o c 2 

1 1 1 1 P 1 5 






i- £ 






v-i 


sjs^-sfii 


— 


_ 


r»"i OS 






C4 




O rt 


© *»■ 


00 t^ 








■ 


i 1 


* 

cs 




, 


4 


,1 

s 


1 

09 


I § 

£ S 


I 

*<> 

« 


■ 

I 


f mi 


!j 1 

8 8.1 

2 t« 2 


ilk 

2 &o a. 



T3 
U 

o 

•mmM 

-o 
_g 

Q 

an 

tU 

l— 

co "d 

5 3 
eo jo 

U 3 



c 


Q 










B 


5 


O 


c 


re 


03 

5 


0/) 




<U 




>■ 


u 


on 


4J 


d 


OJ 




Q 


T3 


C 


C 


C3 


CO 


c 

C3 



« a p 



- 3 

.2 o 
P SP 

DO 3 

"§-1 

« P 

C/5 3 

2 "3 

•r- C3 
.3 O 

2 g 

CO % 

"O v 
>> U 

o ;s 

El '55 
. P 

;c •* 
'3 c 
£ § 

.£ I 
i"g 

a a 

oj C 
'C u 
'e3 c 

P a 

5 *- 

C/l l- 

c/3 O 

5 o. 

co £ 
> 



CO 



u 



0) ^ 

w JO 
^^ ^ ■*-» 

w O 

CO 

£ r 

2 1 
tea o 

« 3 

B IB 

I 8 

x> u 

te <T! 

^ o 

* (U 

^ c 

c o 

•£ E 
£ « 

o ■ 

«J € 
CO (U 

s a 

6 & 

c« 3 



co •£ 

B 3 

ii 

. a « 

co x e 

S B e0 

o "S ^ 
"aft 
oo c ,, 

•£ 2 -= 

CO Ct_ ** 

S g jc 

3 £ O 

"Ob 

9 59 m 

U _ 1) 

* i - 

Cq ™ 

CO CO o 

.-3 "eo co 
S u. « 

•o o £ 

"O <-> 9 

<u o OT 

' O S3 



S.SJ 

on < 



CO 



ir'ji 

°. a> 6 

* < cB 







719 




Table 7-2. Vegetation species absent from the Okefenokee Swamp seed bank samples, 
but present in plots of established vegetation. 








Area Where Species 
is Most Abundant in 


Areas Where Species is 






Species 


Standing Vegetation 

Along Sample 

Transects 


Absent from Standing 

Vegetation Along Sample 

Transects 




Acer rubrum 


Sill Area 


Chesser, Durdin, Sapling 
Prairies 




Brasenia schreberi 


Durdin Prairie 


Chesser, Floyd's, Sapling 
Prairies, Sill Area 






Calapogon sp. 


Durdin Prairie 


Chesser, Floyd's, Sapling 
Prairies, Sill Area 






Carex glomeratus 


Sill Area 


Chesser, Floyd's, Sapling 
Durdin Prairies 






Cephalanthus occidentalis 


Sill Area 


Chesser, Durdin, Sapling 
Prairies 






Cliftonia monophylla 


Durdin Prairie 


Chesser, Floyd's, Sapling 
Prairies, Sill Area 






Fraxinus caroliniana 


Sill Area 


Chesser, Floyd's, Sapling 
Durdin Prairies 






Ilex coriacea 


Durdin Prairie 


Chesser, Floyd's, sapling 
Prairies, Sill Area 






Ilex myrtifolia 


Sill Area 


Chesser, Durdin, Sapling, 
Floyd's Prairies 






Lyonia lugustrina 


Sill Area 


Chesser, Floyd's, Sapling 
Durdin Prairies 






Magnolia virginiana 


Sill Area 


Chesser, Durdin, Sapling 
Prairies 






Myrica cerifera 


Durdin Prairie 


Chesser, Floyd's, Sapling 
Prairies, Sill Area 






Nyssa ogeechee 


Sill Area 


Chesser, Floyd's, Sapling 
Durdin Prairies 





Table 7-2 -continued 









Species 


Area Where Species 

is Most Abundant in 

Standing Vegetation 

Along Sample 

Transects 


Areas Where Species is 

Absent from Standing 

Vegetation Along Sample 

Transects 


Pierus phillyreifolia 


Floyd's Prairie 


Chesser Prairie 


Finns elliottii 


Durdin Prairie 


Chesser, Floyd's, Sapling 
Prairies, Sill Area 


Rhynchospora chalerocephala 


Durdin Prairie 


Sapling Prairie, Sill Area 


Vaccinium corymbosum 


Chesser Prairie 


Durdin, Floyd's, Sapling 
Prairies 


Woodwardia virginica 


Chesser Prairie 


in all areas 












721 



30 
25 - 
20 - 
15 - 
10 - 

5 





u> 15 

© 

o 
i 

w 10 



I 

1 5 

3 

z 



15 



10 



5 - 



n 



n 



4 



Herbs 



n 




I 



u 



L 



Shrubs 



t 



llhi.. 



L 



I I I 



i I i 



=F=F 




Trees 




lllllillii 



~i i i i i i i — i — i — i — i — i — i — i — r~ 

shrubs trees hershr herpra hertre aquher aqushr aqutre 

Structural Zone 
(Increasing Mean Depth) 



Figure 7-3. Counts of species and germinated seeds in seed bank samples, compared to 
species counts in the standing vegetation within the structural zone at the collection site. 
Structural zones are described in Table 6-1. 



722 






lead jo zui jad sBujipaas snoaoeqjan jo jaquinfM aBejeAy 

o 
o 
o 



s 



s 




I s 

£ 

D 

w 

3 

£ 
CO 

1 



(0 
JQ 

L. 

o 

X 



* CO CM t- O 

»eed 3ui jad sBujipaas as-"! Jo qruqs }o jeqwriN e8ej9AV 



J 



5 







723 


Table 7-3. Average number of species in the established vegetation and in the seed bank 
from structural zones throughout Okefenokee Swamp. 


Area, Zone Type, 
and Sample Size 


Average and 

Range of # 

Species 


Average 
and Range 
of # Species 


Average and 

Range of # 

Species in 

Established 






in Established 
Vegetation 


in Seed 
Bank 


Vegetation 
and also in 
Seed Bank 




Chesser Prairie 








aquatic-herbaceous prairie (6) 


5.5 (3-10) 


10.8(9-15) 


3.0 (1-6) 




aquatic prairie (8) 


4.1 (3-6) 


9.4 (4-13) 


2.3 (2-5) 




herbaceous prairie (8) 


8.4 (3-14) 


13.5(10-17) 


4.0 (2-6) 




shrubs-herbaceous prairie (8) 


8.5(4-13) 


12.6(8-16) 


3.6 (2-6) 




shrubs (6) 


8.8(5-14) 


12.3 (9-14) 


3.3 (0-7) 




Durdin Prairie 










aquatic-herbaceous prairie (6) 


6.8 (4-9) 


11.5(7-16) 


4.2 (2-6) 




aquatic prairie (3) 


5.3 (5-6) 


8.3(7-11) 


2.7 (2-3) 




herbaceous prairie (10) 


13.0(7-17) 


14.8(12-18) 


6.7 (4-10) 




shrubs-herbaceous prairie (13) 


16.1(7-24) 


14.0(11-18) 


7.2(3-11) 




shrubs-trees (1) 


22.0 


15.0 


8.0 




shrubs (6) 


16.8(11-21) 


15.3(12-21) 


6.5 (4-9) 





Table 7-3 --continued. 



724 



Area, Zone Type, 
and Sample Size 



Floyd's Prairie 

aquatic-herbaceous prairie (5) 

aquatic prairie (5) 
herbaceous prairie (2) 
herbaceous prairie-trees (4) 



herbaceous prairie-trees- 
shrubs (2) 

shrubs-herbaceous prairie (9) 



shrubs-trees (3) 
shrubs (4) 
trees-shrubs (1) 

Sapling Prairie 

aquatic-herbaceous prairie (1) 

aquatic prairie (6) 
herbaceous prairie (10) 



Average and 

Range of # 

Species 

in Established 
Vegetation 



4.6 (3-7) 



4.6 (3-8) 



8.0 (7-9) 



7.8(5-10) 



10.0(9-11) 



10.1(6-14) 



11.3(10-12) 



12.0(10-14) 



15.0 



7.0 



5.2 (4-7) 



7.0(5-10) 



Average 
and Range 
of # Species 

in Seed 
Bank 



11.6(10-14) 



10.0(7-12) 



11.0(8-14) 



10.8(9-13) 



12.5(11-14) 



10.2(7-13) 



8.7(6-12) 



10.0(9-12) 



16.0 



9.0 



9.7(7-11) 



11.9(9-15) 



Average and 
Range of # 
Species in 
Established 
Vegetation 
and also in 
Seed Bank 



3.4 (3-4) 
3.2 (2-4) 
3.0 (2-4) 
2.8 (2-4) 

4.0 (4) 
3.1(1-5) 
2.7 (2-4) 

2.5 (2-3) 
6.0 



3.0 



3.5 (3-4) 



4.3 (2-7) 



Table 7-3--continued. 



725 



Area, Zone Type, 
and Sample Size 



Average and 

Range of # 

Species 

in Established 
Vegetation 



Average 
and Range 
of # Species 

in Seed 
Bank 



Average and 
Range of # 
Species in 
Established 
Vegetation 
and also in 
Seed Bank 



herbaceous prairie-trees (1) 



herbaceous prairie-trees- 
shrubs (1) 

shrubs-herbaceous prairie (7) 



shrubs-trees (4) 
shrubs (3) 

Sill Area 

aquatic prairie (5) 

shrubs-aquatic prairie (5) 
shrubs-herbaceous prairie (1) 
shrubs (3) 

trees-aquatic prairie (2) 
trees (3) 
trees-shrubs (4) 



10.0 



11.0 



9.1(5-11) 



9.8(8-11) 



9.7(7-13) 



4.4 (2-6) 



6.6 (4-10) 



8.0 



7.3 (5-10) 



5.0 (4-6) 



5.7 (4-7) 



8.3 (7-9) 



12.0 



12.0 



11.1(8-15) 



12.5(9-17) 



10.3(8-15) 



8.4(5-11) 



10.2 (7-12) 



11.0 



9.7(9-11) 



8.0 (7-9) 



8.3 (7-10) 



9.8(6-13) 



5.0 



7.0 



4.3 (2-6) 



2.0(1-4) 



2.7(1-4) 



2.4 (2-3) 



3.8 (3-5) 



3.0 



3.0 (3) 



3.0 (2-4) 



2.3(1-4) 



3.0 (2-4) 



726 



<D 


e 


<u 


0> 


O 


£ 




u 




•*-• 


^r 


TD 


r^- 


C 




<Ti 


u 




3 


21 


H 


$ 



— 

is 




*i 




if If J 
! llf I 



sab; 

5 Cu 



lfj* 



So. 



■ 1 t < 1 
llllf 

all II 



Iff. 

I] ' 



I 

P -is. 
£ 1 * 



■ 3 £ ■ 



f I A 



•I 



N l 






£ 



A 
K 

A 

g 



1 

2 

A 
Q 



- 

V 



K 

* 



I 



If 



if • 
I a 



| 

i 



ri I 

li 



I 

i 



o o 






— — r- r- 

§«*j \o — ■ 

r- so * 

© © do 



ui tt 



s s 



I 



I 






[j 

□ c 



» t s 

ill 



51 



727 




I 



B 






a 

t 



I 



ill 

iff 

I CI 



I 




I I 



I 



i 



I J 



M I: 



If If | 

IJH1 



i 



| 



HE 

III 



III 
III 



Sab 

S a. 



8S 

© r- 

d d 



d d 



— n 

© M 

O NO 

O «r> 

d d 



— &, 

o a 

O ff* 

o r- 

d d 



Hi* 

s S ~ a. 



i |a. 



it 



ill 

UK 



^ w v3 




u 

3 
S 

c 
o 






1M 

■ 3 ■ 



III 






■£ I 









E 
I 



li 



3 § 



728 



III I 
1*1 I 



a 
I 



2 



H 

£ 






8 



8 

a 



9 



111. 
1**1 

ill 



I I 

I 1 



II 




Hill 



> - 

I!: 



i 



M 
II. 



f* 



ffl 



II 



IE 



| 

s 
i 






Err 



jf If J 
fit]] 



■ fl 



I J 



ii 



Et 1 I 



1 § 



I 



Sa& 

S 0. 



S3 

© F 

© © 



© « 








— a 
© n 


B oo 


O 00 


o o 


o o 



III* 

£ ■ 



? I A 



lUil 



u u 



11 



is 

O 6 



T3 

3 
_C 

C 

o 



■a 

H 




ilr 



j: 



< B. 



E g 

vl 



II 



u 



it 
1 1 

K S 
ft; ,3 



J s 

11 



I! 



729 



1 = 1 a 



I 



r* 


u 


o 


£ 


ft 


* 


P 


£ 


A 


u 


o 


a 



9 

A 

Q 




1 
M 



.5 



Mill 



iiiii 

£ x -b J3 i; 



1 

is 



i 



•s £ •= 



I 



If ii 
Si fj 



|g 



i 



] a a 
5 a. 



o — o 



82 

O OC 

© © 



If 1 A 



-a 

3 
fl 

B 

O 
o 



1 




Hi* 

in* 



- E 



I I 

Ikf 



ill 



I 



1 



! 



'So a. 
5 >-> 

f I 






730 



In I 



i 



811 

» * < 



a 



a 






! 



I 



"5 
J 



t 



IllII 
11*11 



I 

s 

■ 



fill 
I til 






«; b 



ft |1l lie 

•5 « iili -Si 



I 



if if I 

1J«1 



Hi 



■1 



ill 



i 



II 



k'J 



I ft, 






1c I 1 

Si 13 ' 



S 






Hill 



- ** 

oil oil 

I ■ B c to v 



H M H 



111! 



i I 

d 6 



i i: 



1> 

3 
S 

c 
o 

4 

1 
H 



ii i 



i 



It! 



1 6 
fj 

> S 



oSi 






111 

X > Si 



5 
1 

f is 



'53 

2 > ■ 



Vt 

co 

-o 

<u 

+-» 
CO 

'> 
u 

c 

1 

g 

(S3 

>> 

p 



o 

o 

"o 

• o 

© I 

<u e 

VI — 

O Q. 

x c 
2 5 

•— CO 

II 

<U CD 
"C * 

u 



731 



Vl 

X> 



OS t- 



X 

en 
i 

<D 



CO 
(U 



VI 

T3 

c 

CO 



B, S 
o °- 

a 5 
cr K 

« 2 

-^ " 
L- £ 

■S J 

* -5 
<-< f 

If _c 

I C/l 

ID CD 

I g 



a -a 



a 



<u 



u. x 

<D S 

i— 

CO 19 



a 
a 


O. 


cr 

C3 


v> 

3 
O 


u 


cj 

o 




CO 


"3 


X) 


cr 


u. 


C3 


CJ 





A § 




a- ■£ 




. o 




"O _ 




C cO 




« c 




_r O 




T3 -S 




1 | 




5J CD 




T3 H 




O g 




£ V5 




i- o 




CO OT 




v> §j 




<-> c 




<D & 




D. ?£> 




to U 




ci 




£P- 

c c 




'C <D 
C en 




3 O 




5 o 




o g 




>~> c« 


1 


monl 

type 


C 


E g 


c 


S a 


o 


O N 




Only most 
Structural 


1 


H 


a JD 



-a i 

C/5 -^v 

3 £ 

° -s 

8 3 

p cfl 

E .2 



'C 

J 

a. 

00 



ex 

CO 

a. 
tvo 

c 

'5 

k- 

a. 
</i 

•b 
>> 
o 



'Cj 

J 

a. 

c 

•B 

i- 

3 

a 



cj co 

■ — — 

§ B. 
3 «q 

4S 8 

h 9 

cj ■ 

II 

-— VI 
b. u. 

'3 <" 

S JB 

rs CO 

cr g, 

S » 

2 § 

. x: 



co 



.2 

c/i 'C 

E 2 

ps ^ 

J X 



U 



•S u 



x 



c 2 



C3 



V3 

I e- 

o y 



IS « 

x x 

co CO 



CO 

CO 

1) 









732 
(Table 7-4). Lowest diversity occurred in Floyd's Prairie samples in the autumn and 
Floyd's and Sapling Prairies samples and the Sill area samples in the spring. Chesser 
Prairie samples consistently had the highest diversity of species, whereas Durdin and 
Floyd's Prairies samples had the greatest number of germinated seeds. The high 
germination rates in samples from these areas were due primarily to the abundance of 
yellow-eyed grass, water lily {Nymphaea odorata), and beakrush {Rhynchospora 
inundata) in Durdin Prairie samples, in addition to redroot (Lacnanthes caroliniana) in 
Floyd's Prairie samples. Species numbers per sample were more variable in the autumn 
than in the spring, although overall means did not differ (Table 7-1). Total counts and 
species diversity were greater for the exposed treatment, whereas species diversity was 
more variable among samples in submerged treatments (Table 7-1). 
Trends in Response to Hydrologic Conditions 

Responses of seed bank seedlings to gradients of water depths mirror those of the 
standing vegetation (see Chapter 6). Species in the seed bank could be loosely grouped 
into associations identified in Chapter 6, based on a general gradient of average water 
depth and variability (Table 7-5). Along this gradient, species in the seed bank that were 
found as standing vegetation in constant, deep water conditions (see Chapter 6) were less 
variable in numbers in the spring than in the autumn (Table 7-6). Many of these species 
(e.g., broomsedge, Andropogon virginiana; creeping rush, Juncus repens; bamboo 
greenbriar, Smilax laurifolia; Walter's greenbriar, S. walteri; red root; spikerush, 
Eleocharis robbinsii; dahoon holly, Ilex cassine) are autumn seed producers, which 



733 



Table 7-5. Hydrologic environments where seed bank species are found in Okefenokee 
Swamp, and areas of maximum species abundances in seed bank samples and established 
vegetation. 



Species 


Area Where 

Greatest 

Abundance 

Occurs in Seed 

Bank 


Area Where 
Greatest 

Abundance 
Occurs in 

Established 

Vegetation 


Hydrologic 
Zone Type 
Where Most 
Frequently 
Found in 
Established 
Vegetation 


Structural Zone 

Type Where Most 

Frequently Found 

in Seed Bank 


Andropogon sp./ 
Erianthus sp. 


Durdin Prairie 


Durdin Prairie 


constant, deep 


shrubs-herbaceous 
prairie 


Bidens mitis 


Durdin Prairie 


Durdin Praine 


constant, 
moderately deep 


shrubs-herbaceous 
prairie 


Carex walteriana 


Floyd's Praine 


Floyd's, 
Sapling Prairies 


moderately 
variable, deep 


herbaceous prairie- 
trees-shrubs 


Clethra alnifolia 


Sill Area 


Floyd's Praine 


moderately 
variable, deep 


shrubs 


Cyperus erythrorhizos 


Sill Area 


Sill Area 


moderately 
variable, deep 


trees-shrubs 


Cyrilla racemiflora 


Sill Area 


Chesser, 
Sapling Prairies 


moderately 
variable, shallow 


trees-shrubs 


Decodon verticillatus 


Sill Area 


Durdin Prairie 


moderately 
variable, deep 


aquatic prairie- 
shrubs, 
trees-shrubs 


Drosera intermedia 


Sapling Prairie 


Durdin Prairie 


constant, deep 


shrubs-trees 


Dulichium 
arendinaceum 


Chesser Prairie 


Chesser Prairie 


variable, deep 


aquatic prairie-trees 


Eleocharis 
baldwinii'vivipara 


Sill Area 


Sill Area 


constant, 
moderately deep 


trees 


Eleocharis robbinsii 


Floyd's Prairie 


Floyd's Prairie 


constant, deep 


aquatic praine 


Erianthus giganteus 


Durdin Prairie 


Durdin Prairie 


constant, deep 


trees 


Gordonia lasianthus 


Sill Area 


Chesser Prairie 


constant, shallow 


aquatic prairie- 
shrubs, herbaceous 
prairie-trees 


Ilex cassine 


Floyd's Prairie 


Floyd's, 
Sapling Prairies 


moderately 
variable, shallow 


trees-shrubs 


Iris virginiana 


Chesser Praine 


Floyd's Prairie 


constant, deep 


herbaceous prairie- 
trees 



Table 7-5-continued. 



Species 



Ilea virginiana 
Juncus repens 

Juncus triganocarpus 



Lacnanthes 
caroliniana 

Leucothoe racemosa 



Ludwigia alata 
Lyonia lucida 

Lyonia sp. 

Nuphar luteum 
Nymphaea odorata 



Nyssa sylvatica v. 
biflora 

Orontium aquaticum 



Panicum hemitomon' 
Sacciolepis striata 

Rhynchospora alba 



Rhynchospora 

cephalanthaf 

microcephala 

Rhynchospora 

fascicularis/ 

wrightiana 



Area Where 

Greatest 

Abundance 

Occurs in Seed 

Bank 



Sill Area 
Sill Area 

Sill Area 

Floyd's Prairie 

Durdin Prairie 

Floyd's Prairie 
Chesser Prairie 

Sill Area 

Sill Area 
Durdin Prairie 

Sill Area 

Floyd's, 
Sapling Prairies 

Sill Area 
Durdin Prairie 
Durdin Prairie 

Durdin Prairie 



Area Where 
Greatest 

Abundance 
Occurs in 

Established 

Vegetation 



Floyd's Prairie 
Sill Area 

Sill Area 

Durdin Prairie 

Sill Area 

Sill Area 

Chesser, Durdin 
Prairies 



Sill Area 
Chesser Prairie 

Sill Area 

Chesser Prairie 



Sapling Prairie, 
Sill Area 

Durdin Prairie 



Floyd's Prairie 



Durdin Prairie 



734 


Hydrologic 

Zone Type 

Where Most 


Structural Zone 




Frequently 

Found in 

Established 


Type Where Most 

Frequently Found 

in Seed Bank 




Vegetation 






variable, deep 


shrubs 


moderately 
variable, deep 


trees 




constant, shallow 


aquatic prairie- 
shrubs 




moderately 
variable, shallow 


herbaceous prairie- 
shrubs 




moderately 
variable, shallow 


shrubs 




constant, deep 


shrubs-trees 




moderately 
variable, shallow 


shrubs 




moderately 
variable, shallow 


herbaceous prairie- 
trees 




variable, deep 


aquatic prairie-trees 




moderately 
variable, deep 


aquatic prairie 




variable, deep 


aquatic prairie- 
shrubs 




moderately 
variable, deep 


aquatic prairie 




moderately 
variable, deep 


trees-shrubs 




moderately 
variable, shallow 


aquatic-herbaceous 
prairie 




moderately 
variable, shallow 


herbaceous prairie- 
shrubs 




moderately 
variable, shallow 


shrubs 





Table 7-5--continued. 



735 



Species 



Area Where 

Greatest 

Abundance 

Occurs in Seed 

Bank 



Area Where 
Greatest 

Abundance 
Occurs in 

Established 

Vegetation 



Hydrologic 
Zone Type 
Where Most 
Frequently 
Found in 
Established 
Vegetation 



Structural Zone 

Type Where Most 

Frequently Found 

in Seed Bank 



Rhynchospora 
inundata 

Rhynchospora spp. 



Saggetaria graminea 



Sarracenia flaw 



Sarracenia 
psittacenia/flava 

Scleria reticularis 



Smilax laurifolia 
Smilax walteri 



Smilax spp. 



Syngonanthus sp./ 
Ericaulon sp. 

Syngonanthus 
flavidulus 

Taxodium ascendens 



Triadenum virginicum 

Websteria sp. 
Xyris sp. 



Floyd's Prairie 



Durdin Prairie 



Chesser Prairie 



Durdin Prairie 



Durdin Prairie 



Floyd's Prairie 



Durdin Prairie Durdin Prairie 



Durdin Prairie 



Chesser Prairie 



Chesser, Durdin 
Prairies 

Durdin Prairie 



Chesser Prairie 

Durdin Prairie 

Durdin Prairie 

Sapling Prairie 

Chesser Prairie 

Durdin Prairie 
Durdin Prairie 



Durdin Prairie 



Chesser Prairie 



Durdin Prairie 



Floyd's Prairie 



Durdin Prairie 

Durdin Prairie 

Sapling Prairie 

Durdin Prairie 

Durdin Prairie 
Durdin Prairie 



moderately 
variable, deep 

moderately 
variable, shallow 

constant, 
moderately deep 

constant, 
moderately deep 

constant, 
moderately deep 

constant, shallow 



moderately 
variable, shallow 

moderately 
variable, shallow 

moderately 
variable, shallow 

constant, deep 



constant, deep 

moderately 
variable, deep 

constant, 
moderately deep 

constant, deep 

moderately 
variable, deep 



herbaceous prairie- 
trees 

herbaceous prairie- 
trees-shrubs 

herbaceous prairie 



shrubs-trees 



shrubs 



herbaceous prairie- 
trees-shrubs 

shrubs 



trees-shrubs 

herbaceous prairie 

shrubs-herbaceous 
prairie 

shrubs-herbaceous 
prairie 

herbaceous praine 
trees 

aquatic prairie 

shrubs-herbaceous 
prairie 



736 



Table 7-6. Effects of season on response of seed bank samples (counts) collected from 
hydrologic zone types and areas of Okefenokee Swamp. 





Seasonal 








Hydrologic 


Variances 


Season with 


Seasonal 


Season with 


Zone Type or 


Differ, 


Larger 


Means Differ, 


Larger Mean 


Area 


P>F 


Variance 


P>t 




Constant, Deep 


Yes 


Autumn 


No 




Water 


O.OOOl 








Constant, 










Moderately 


No 




No 




Deep Water 










Moderately 










Variable, Deep 


No 




No 




Water 










Variable, Deep 


Yes 


Autumn 


Yes 


Autumn 


Water 


<0.0001 




0.0001 




Moderately 










Variable, 


Yes 


Autumn 


Yes 


Autumn 


Moderately 


0.0021 




0.0001 




Deep Water 










Constant, 


Yes 


Autumn 


No 




Shallow Water 


<0.0001 









737 
probably contributed to this trend (Table 7-7). Differences, by structural zone, in 
seedling numbers were not significant in this species' group, although treatment 
responses differed (Table 7-4). Highest germination occurred in exposed treatments 
(Table 7-4). As water depths decreased but variability remained constant, structural zone 
differences continued to be non-significant, which probably relates to wind as the 
predominant dispersal method of these species (Table 7-7). Differences between 
treatments declined along this gradient of water depth. With gradually decreasing water 
depths, variability in seedling counts continued to be higher in the autumn (Table 7-6). 
The low variability in water levels where these species occur across the depth gradient, 
and the ability of many of these species (e.g., broomsedge, redroot, yellow-eyed grass, 
spikerush) to germinate in exposed and submerged conditions, suggests generalist habits 
that result in high seed and seedling survival of these primarily wind-dispersed species 
throughout a broad range of water depths. 

Species in the seed bank that were found as standing vegetation in more variable 
water depth conditions (see Chapter 6) showed more pronounced seasonal changes in 
abundance and variability in the seed bank as water depths and variability increased 
(Table 7-4). Differences in seed bank composition among structural zone types were 
also more apparent in these species (Table 7-3). Structural zones with the lowest density 
of germinated seeds could be grouped into 2 types. Aquatic and herbaceous prairie 
structural zone types generally occurred where water levels were fairly constant; 
therefore, abundances of species found more often in variable environments were not 
expected in these structural zones, as illustrated in Table 7-3. Samples from structural 



738 



■ 
1 a 

© c 

a ? 

I ! 



I 

f| 
| J 

3 

e 



8 8 



1 

1) 
■ 



B 

■c 
I 

a, 



I 

E 
■ 

i 

w a; 
C 
I 



-O 

E 






I 

5 

1 

* » 

S3 x> 

U c/) 

-S <D 

B £ 

J3 a 



3 
| 



Crt 

2 
8 



e 

■e 

'S 



J 



£ 



3 E S 

w tn -C 



o 

I 

E 



i 

■c 

a. 
u 

I 

3 

I 



£ 



t% s 

« «, L, 

s i 1 

•C 5 .2 

a. S q 



-a 

s 



u 


ID 






CO 


es 


is 


•5 







IU 


V 






t3 


■5 


* 


■5 






l/l 



73 

E 



E 

3 



E 

3 

o 

E 



73 
E 



E 

3 

<u 

E 



"3 
E 



73 

E 



73 
E 



73 
E 



a> 




<i> 






c 


OC 
ON 


cJS 


^H 


<L> 


s »— ^ 


s 


u. 


O 


O 

JZ 


C4— 1 


-*-* 





1 


c/5 











T3 

C 




« 






O 


C 
O 


e3 

J3 


U 





T5 




C 


as 


cC 


C/l 




u- 


/■"•-«, 


k 


ON 




ON 


T3 


1—1 






■O 




C 


1> 


CO 


X! 


c 


H 








13 


CL, 


c 


1/1 


£ 


"a. 


<D 




O 




crt 


r>- 


^ 


1 


C 


r^ 


CO 





X) 


H 


T3 

SJ 
t/2 



■all 

2 - c 

Ifl 



IU 

E 
E 

3 

in 
<u 
I 



2 
3 



E 
E 

3 

CO 


hi 
IU 

E 

£ 

3 

VI 


>* 


IU 


fa 


ts 


IU 


— 1 



I 

6 
I 

£ 

< 

a. v> 

E t 

k a. 






S i J ?. I 

.2 « I I eo 

» " « « _ 

9 w 



a 

(A 



§ 



-o 

E 
3 = 

Q £ 

I" J 

if 



•a 
I 



c 

1 
O 



1 

.5 
•p 



M GO 

fe op 

vt is 

J 1? 



W3 






-a 

u. 

.S 
-a 

3 

a 



o 
U 



■a 
I 



c 



3 


— 


a 


C/3 


u. 


00 


CA 


a 


V) 




<L> 


n 


£ 


in 





c^ 



c 


B 


c 

(D 


























B 

a. 


■ 

a. 


J= 



g 


§5 


V) 


(A 


g 


i 


a. 


■b 



i 



3 v> 

II 



•« 



I 



I 

•Si 
1 

I 



c3 



& 




s- 


R 






« 


•4S 




*> >» 


It! 


e 2 


•Si 

a 


Cype 
eryth 



2 
I 1 



-a 
I 



.S 
■a 



53 op 

VI c 

d a. 

U W3 



01 

c 

E 

OS 

1 1 

U c« 



IU 

o. 





5 




5 


1 


| 


2 

6^ 




g 

5y 


c 


2 


Si 


Q 


> 


q 


c 



II 

q S 



■a 



c 

1 

3 S3 

Qifl 

53 gp 

c/i « 
I & 
U c/i 



c 



2 
.1 



i 



a 



739 



ev ■*■- 

E£ fJl 

CO S * 



a 2 E 

j it 

- £ .22 

a. 2 £ 



J 

co 

■o 

SI 
co 



ill 
if I 



- j 

I e 

S 1 

£ ee 

so T3 

II 

M "~ 

1 | 

E t 

« a. 

CO 



<L> 
■g 

c3 
a. 
o 

w 

3 

7 



cu 

£ 



S >- 




x> 


en 




■- a> 

D.X 
<2 en 


to 


E 


3 

5 




CU 


CO 


o 




1 3 
3 2 


— 

■a 


to 
B 


83 

•fi 


B 


ca w 


n. 


to 


CD 


i- 





B 




■g 


■ 


'a 


g 


E 


•g 


Q. 


'1 


en 


E 


3 


a. 


8 


.a en 


U en 


I -3 
3 g 


eS X) 


| 5 


?•§ 


J> J3 


cs en 


JS en 



X) 

3 

i 



ce) 

I 

C 



c 



11 


<o 






<a 


B) 


i 


* 



-a 
E 



73 

E 



13 
E 



eu = 



3 



3 

9 

8 



o 

3 
C 

■♦-» 

c 
o 
o 






^* 5 en l. 9 
tn " ■ ■ _ 



"S 

CO 



i 



5 = 
fig 

in C 

en ls 

cu Q. 
U C/3 



-a 
e" 

E 

3 S3 

fe op 

en C 

en is 

J I 

U co 



00 

1 

c« 



•a 



eu 

D 



g 



eu 

a. 



jj 

-c 



I 3 

"5 -Si 

§ I 

SS Sb 



•S S 

1 -s 

•5 1 

o S 



i 



s 
J 



1 



4 

c 



5 



I 



5 

c 



I 

c 



E E E 

tn en en 



•a 

E 



13 

E 



cu 

E 
E 

3 
in 

eu 

13 



s 



2 

3 
CS 



1) 

E 
E 

3 



I 



Cfl 



e/5 



cu 

u 



1 

3 

a 



-o 



.S 

T3 



S3 3 

£ 



c 5 



eu 30 



C en C 

— en ^s 



I 

00 



en ^s 

cu Q. 

A s 

U co 



*— 

c 


| 


B 


S 


cu 


in 




CO 


en 




i 


E 

cu 

a. 


g 

B 
a. 


1 

cu 

a. 


I 



f 



1 I 



8 



y 

5 



If 



2 8 
1 j 

SI 

i i 

5 8 



eu 

© 

-5 

I 

eu 



cu 

s 

5 



-o 

cu 

E 



5 

3 

co 



u. 



3 =3 

fia 

cu 00 

en C 

cS "a. 
U co 



g 



CU 

a 



9 
Jj 

a 

:& 
I 

>5 



740 



£ ■ £ 



£ 4/ 



1 * 



Vj 



V) 



121 



cu 

a 



V) 


u 




-C 


C(3 




R 


X) 


u 








■ 


_E 


6 



CU 

•c 

■a 

E 

a. 
u 

■•g 

ctj vi 
3 0) 

cii. 



a 


V 


CO 


•B 


■c 


■s 


'SI 


1 


a 


c 


E 


u. 


a. 


O. 


a 


u 


u CO 


o 


I 


11 

3 £ 


nj 


3 


3 


7 


TO CO 


1? 



CO 
cu 

±3 





u 


CU 




c 


■B 




•a 


1 




L_ 


E 




O. 


a 




3 


s 


1 


8 

CJ 


O c/l 




rt 


P3 .o 


3 


Xi 

E 


1 1 


I 


■C 


J= VI 



bo 8 

5 s - 
s 1 1 

'C .2 .2 

a. 2 q 



1> 


■ 


■ 


B 










£ 


« 


o) 


CO 


* 


* 


* 


■s 



T3 

c 



en 
1 



1 
E 



73 

E 



a 


V 


■ 


B 


60 


oo 


00 


00 


g 


1 


§ 


5 



73 
E 



73 
E 



73 
E 

co 



. I ■ 
1 £ © 
I i | 

"5* 



■ 

E 
E 



D 

E 
£ 



Cu 

E 
E 

3 



1 S 
cn "O 

n 

E I 



- I .5 4 

■S * '5 r j 

1 « CO 



s 
a 



<u 

3 

C 
• »■« 

*-> 

e 
o 
u 



X> 



Is»i 

■" w «. V t» 



"S 

=. 
to 



•a 



53 op 

co C 

co ^s 

I I 

U c/j 



C/3 

SO 

B 

ft 



O 



-a 



3 =3 

fc" op 

CO C 

CO _3 

1 1 

U en 



§ 



.s 

-a 

3 = 

«3 SP 

CO C 

<G "5. 
rs * 

U C/3 



I 

a- 



00 


CO 


fi 


73 


| 


| 


c*l 


u- 


CO 


|M 

0) 


T3 




| 





co 



u. 



c 

E 
3 

Q 



.c 
U 



c 



CU 
Cl 



cu 

Q. 



g 



E 

Q. 



£2 

cu 

Q. 



■8 



5 
I 



.a 

I 



i 

6. a 
ST »■ 



si 






S S I 

s « a 






S 
1 

SI 
1 i 



a, 4 =o 



CJ 

I 






3 

a 


c^ 


ii) 


on 


in 


H 


V) 




B 


D. 




I 


u 


c^ 



c 
B 



£2 

cu 
Q. 



2 



| HI 



14 
It 



741 



C 



2£ 



m 



1 2 E 

ill 

ft. S Q 



•c 



u 

OS 



to u 



•a 

1 

8 J 

U CO 

* in 

-£ <u 

B g 



t> 




w JD 


£ 

i 




8| 


to 


V) 


o eO 


X) 


X) 


i ,1, 


5 


3 


■£ <u 


JC 


JC 


u 2 


en 


(A 


JC -E 



s -s 





■fi 


3 
g 




E 


U 




a. 


rt 


5 


3 

8 


X> 

u. 

a> 

JC 


(O 


9 






■ 


,.n 


a> 


€ 


? 


s 


■ 

JC 


JC 



to 



E 


■ 


V 


E 


a 


B 


(U 


ID 


B 


S 


1 


S 


oS 


« 


13 


* 


2 


S 


* 


5 


* 


S 


5 


is 



Cfl 






— a. i 
•z & * 



1 

I 

< 

m 

c 



Jf 

B 
« 

t 



a. 3> 

E £ 

~ ft. 



1 



£ 

3 

E 



73 

E 



CO 

E 



1 

E 



E 

3 

s 

V 

E 









60 2 

S E 

— to 



(LI 

E 
E 

3 



B 

E 
E 

3 
en 



E 
E 

3 



E 

E 



n 
E 
E 

3 



3 2 

3 3 

CO CO 



•o 

3 
_C 

•+■» 

c 
o 
o 

I 
I 

I 

1) 

1 




a 

t/3 



•a 

u. 

s£ 
xi 

c 
3 — 

® op 

J I 



•O 

Ei 

1- 

OS 

<U 50 
en C 
en JC 

i I 



I 



C 



fe 60 
en C 

en JC 

JC * 

Jm CO 

U on 



E 
.s" 

-a 

c 

3 

a 

te op 

in C 
en iE 

A r 



c 



c 
s 

3 

a 



t/3 

60 
& 

I 



W3 
60 

I 

I 

cf 

1 

Q 



8 

JC 

D 



C/3 



-a 



c 
u 

3 

Q 



6 

JC 





c 



IU 

JC 





§ 



a. 



a. 



c 

0) 



a> 
a. 



J2 12 

a. a. 



G £ 



2^ 
ll 



I 



5. a 



£ 



*,* 



9 S 



2 
5 

| 

3 

■i 



si J f 



•a s 

<^l 6C 



J 



,3 



| 
.a 



t 



Si 

1 



a. 
o. 



^ J ^ 



I 



I 



I 



E 
E 

3 



60 "1 



° | J? 



c 

3 =3 

Qw 

fe* ep 

en C 
tn a 

I & 

U t/5 



I 1 J I 



iu 
a 



« 5 

ll 







a, >> 


V] 


U 






co 




Structural Zon 

Type Where 

Most Frequent! 

Found in Seed 

Bank 


3 


•c 
'1 

E 
O. 

s 

8 

Q 

■e 

CD 






3 




8 
8 

■e 

B 
dfl 

en qj 
3 '6 


B 

<d 


CD 

•c 

c 
a. 
u 
1 

3 

cr 


8 

B 
■ 

| 

CD 

-C 
I 

co a> 

5 -a 






VI D. 


JC 


■fa 


CO 


co D. 




s * r 


■ 


kg 

H 




I— 

B 


T3 




11 i 


■ 


1 


ts 


os 


C 




•C 5 .2 

ft. 2 Q 


S 


£ 


s 


* 






J 


1 


B 
00 


9 


^ 


^ 




1 


£ 


J 


E 


E 


E 




■ 




co 


CO 


CO 






w. 














B 


1 






1_ 




nitial 

spersal 

eason 


E 

E 

3 

n 


if 

3 C 
« Q. 


B 
E 
E 


B 


B 
E 
E 

3 
CO 




"1" 


■ 


<D " 


3 

CO 


3 

CO 


ffi 

J 




S-* 














1 § 


«f 








co" 




£ ss 


TJ 








T3 




CO "O 

OS «> 

SI 


| 

E 




=3 

o5 




1 

u- 




< = 


jf 




I 




i 




.5 c 


ha 

3 




i 




E 
3 = 




t I 


Q 









a ■<% 




E £ 

re ft. 




j 


a> 

CO 

1 

4= 


C 


B* 06 






1 1 


& 


3 


J 1 






U oo 


C/3 


u 


Q 


O c/) 




*■ 














tent 
ar) 
sien 
r) in 
ank 




C 

■b 


I 


c 
fl 


I 




Persis 

rTran 
<1 yeai 
SeedB 


to 

1 

■ 

O. 


B 

a. 


co 

g 

B 

O. 


CO 

CO 

E 

B 


"O 




C '»' 












d> 
















3 




























































C 
















a 
















o 




<f? 


co 










1 
IN 

u 

3 

03 

H 




■ 

'2 

o. 
oS 


Syngonanthu 
flavidulus 


•S 41 

i 1 


Triadenum 
virginicum 


o. 

CO 

i 


ci. 

CO 

1 



743 
types of shrubs-trees, trees, and aquatic (or deep water)-trees structural zone types also 

had low numbers of germinated seeds (Table 7-3). Although these zones may be found 
where water levels are more variable (e.g., the trees and aquatic-trees zone types were 
found only in the sill area) and would therefore be expected to have greater numbers of 
species that are found in the standing vegetation of hydrologically variable environments, 
these zones had sparse seed banks. The herbaceous understory cover was not dense in 
these areas, most likely due to dense shrub and tree growth, and deep water levels in the 
sill area. Therefore the contribution of herbaceous species to the seed bank of these 
structural zones was small. However, herbaceous species in the standing vegetation that 
were more abundant where water level variability was greater, were also more abundant 
in the seed bank of these zone types (Table 7-3). 

Treatment type also significantly affected the germination response of species 
that most frequently occur under variable hydrologic conditions (Table 7-4). Many of 
these species, particularly woody species that germinated from the sill area seed bank 
samples collected in the spring, disperse their seeds in the autumn and early winter; 
survival of these seeds is probably enhanced if water levels are at their annual low levels 
when this seed rain occurs. Later, rising water levels due to increasing late-winter 
precipitation transport seeds away from the parent plants, possibly distributing them to 
suitable germination sites. 

Seed bank and standing vegetation compositions were most similar where 
vegetation structure was most complex (Table 7-3). Shrubs-aquatic prairie, herbaceous 
prairie, shrubs, shrubs-herbaceous prairie, and herbaceous prairie-trees-shrubs had the 



744 
greatest similarities between the seed pools and established vegetation, and the latter 3 
zones had the greatest species richness in the seed bank and established vegetation. 
These structural zone types also corresponded generally to areas of relatively constant 
water depths, which is reflected in the list of dominant species in the zones (Table 7-5). 

Discussion 

Wetland Seed Bank Composition and Vegetation Community Dynamics 

Wetland seed banks provide clues to historic vegetation (Leek 1989), suggest 
current species dynamics and departures from historic conditions in the environment, and 
indicate potential responses to future environmental variability and disturbances. 
Importance of the seed bank in wetland dynamics varies with individual wetland and 
wetland type, although similar trends in seed bank contents and structure among 
wetlands reflect similar environmental dynamics and their effects. Thompson and Grime 
(1979) identified 4 seed bank strategies that result in seed bank temporal and spatial 
variability (Type I: transient summer and autumn colonizers; Type II: transient winter 
and spring colonizers; Type III: persistent or transient; Type IV: persistent). Species 
representing all of these germination strategies were present in a range of densities in the 
sampled Okefenokee Swamp seed banks (Table 7-7), indicating that a variety of 
responses in seed germination and the established vegetation community that develops is 
possible with environmental variability. 



745 
Transient species, which persist in the seed bank for less than a year after 
dehiscence, are mostly summer and autumn annual and perennial grasses that colonize 
dry or disturbed habitats (Type I), and Type II species or annual and perennial herbs and 
woody species that colonize gaps in late winter and early spring (Thompson and Grime 
1979). Seeds of these species are usually large, readily germinate in light or dark 
conditions, and are generally found near the soil surface (Leek 1989, Thompson and 
Grime 1979). This type is represented by cypress, blackgum (Nyssa sylvatica v. biflora\ 
red maple {Acer rubrum), loblolly bay {Gordonia lasianthus), sweet bay (Magnolia 
virginiana), swamp red bay (Persea palustris), titi, fetterbush (Leucothoe racemosa), 
hurrahbush (Lyonia lucida), Walter's greenbriar (Smilax walteri), and bamboo greenbriar 
(S. laurifolia) and other shrubs and trees in Okefenokee Swamp. Transient herbaceous 
species in the swamp include arum (Peltandra virginica), spatterdock (Nuphar luteum), 
waterlily, goldenclub (Orontium aquaticum), blue flag iris (Iris virginiana) , and narrow 
leaf sagittaria (Sagittana graminea). Persistent species, which remain viable in the seed 
bank for >1 year, may have a transient component in the seed bank (Type III) or are 
completely persistent with a large sub-surface reserve (Type IV) (Thompson and Grime 
1979). These species are usually small-seeded, and require light, alternating 
temperatures, and aerobic conditions to stimulate germination (Leek 1989, Thompson 
and Grime 1979). Approximately 95% of the seeds collected from the Okefenokee 
Swamp seed pool and germinated in exposed or inundated conditions represent the 
persistent component. 



746 
Just as seed bank composition can affect standing vegetation composition with 

changes in the site environment, seasonal and annual dynamics of the standing vegetation 
affected by disturbance (e.g., fire, scouring by flooding, animal activity), disease, and 
hydrologic cycles can significantly affect seed bank composition (Leek 1989). Thus the 
diversity, size, and composition of the seed bank may provide clues to a wetland's 
disturbance, hydrologic, and succession history (Leek 1989). In some wetlands fire is an 
important influence on seed dynamics, whereas it has a minimal effect in others (Smith 
and Kadlec 1985). Light and nutrient availability and moisture conditions are drastically 
altered by peat and surface fires, and response of the vegetation community to these 
changes may be rapid. The seed bank probably plays an integral role in post-fire 
vegetation dynamics of non-woody species in Okefenokee Swamp. Cypert (1973, 1961) 
found that within a few years after fire woody species in Okefenokee Swamp were 
recovering from burn damage predominantly through coppice growth and stump 
sprouting, and except where burns removed peat and killed root systems, composition of 
woody species was approaching that before the burn. Cypert made no tally of seed- 
sprouting woody species; regrowth was primarily through stump sprouting. Herbaceous 
response was also rapid and included a mixture of beakrush and redroot within the first 
post-burn year, and chain fern (Woodwardia virginica), sedges {Car ex spp.), yellow-eyed 
grass, redroot, and bur marigold (Bidens mitis) within 2-3 years (Cypert 1961). Although 
some species replacement occurred, most of the herbaceous species established within 
the first few years were present 15 years later. However, woody species were slowly 
displacing herbaceous growth. Several of the species and trends recorded in the seed 



747 
bank study herein were similar to those recorded in Cypert's post-burn study plots 
(Cypert 1973, 1961). Disparities suggest that post-burn herbaceous response is not 
completely dependent on the seed bank. Walter's sedge {Car ex -waiter iand) and chain 
fern were important species in the initial post-burn recovery in the late 1950s; although 
these species were abundant in this study where seed bank samples were collected, they 
were poorly represented in the seed bank samples. Walter's sedge may recover from 
surface fires that do not burn into the peat and kill the roots, by resprouting rather than 
seed germination, which gives the species a competitive edge over those recovering from 
fire by germination. Chain fern was not recorded in the seed bank samples during the 
experiment interval, but appeared in sample trays that were retained for seedling 
maturation and identification within a year of sample collection. Fern spores are 
probably abundant in the peat samples, and their presence was overlooked due to the 
brevity of the germination study. Fern spores were estimated to be 8-100 times more 
abundant than seeds in Malaysian peat (Wee 1974). Redroot seedlings originated from 
seeds and rhizome segments (3%) in the Okefenokee Swamp peat samples. Sandhill 
cranes (Grus mexicana) graze heavily on redroot shoots and rhizomes in Okefenokee 
Swamp (Cypert 1961); regrowth from rhizome segments may provide more rapid 
recovery from this feeding activity than germination from seeds. Other persistent species 
recorded in abundance in the post-burn plots (e.g., yellow-eyed grass, 3-square, 
Dulichium arendinaceum, beakrush) (Cypert 1973, 1961) were also abundant in the seed 
bank samples in this study. 



748 
Seed bank differences among areas, structural zones, treatment responses, 
seasons, and standing vegetation composition help elucidate current and potential spatial 
diversity in the Okefenokee Swamp vegetation community dynamics. The seed banks 
sampled in this study display characteristics common to other freshwater wetland 
systems. Most of the sampled seed pool was comprised of the same few species 
throughout the swamp. Dominant species observed in this study were similar to those 
recorded in other seed bank studies (Gerritsen and Greening 1989, Conti and Gunther 
1984, Gunther et al. 1984). However, proportions differed as a function of sampling 
technique and emergence methodology, and also due to pre-sampling conditions in the 
swamp. Gerritsen and Greening (1989) sampled after a period of low water and their 
density measurements may have been inflated for drought-response species such as 
beakrush and redroot, while deep-marsh species may have been under-represented. Their 
short study duration (during 1 year) also could not quantify annual seed bank variability 
that would reflect inter-annual environmental variance. 

Over-representation of yellow-eyed grass, 3-square, and redroot in the 
Okefenokee Swamp seed bank may reflect their dispersal mechanism (wind), the large 
potential seed production contributed annually, and the longevity of their seeds in the 
submerged sediments. Monocots such as these are not uncommon dominants in wetland 
seed banks; frequently the dominants in the seed bank are perennials that can produce a 
large annual seed rain in rapid response to environmental variability, and thus perform as 
facultative annuals in an otherwise "annual-poor" environment (Leek 1989). This 
prolific production of seeds results in seed bank persistence that disproportionately 



749 
represents the species in the wetland' s vegetation history. Although these 3 species were 
found in the established vegetation of all sampled areas, they were minor standing 
components of the prairie environment, which comprised <10% of the swamp landscape 
(see Chapter 4). Assessment of standing vegetation while conducting seed bank studies 
is integral to recognizing these disproportions (also see van der Valk and Davis 1979). 

Forest covered nearly 60% of the Okefenokee Swamp landscape, yet woody 
species occurring in these areas accounted for <1% of the germinated seeds. Woody 
seed presence is usually low in wetland seed banks due to low seed production and low 
seed survival in anaerobic conditions. Unsuitable conditions for germination, type and 
state of decaying peat, patterns of standing vegetation (which are also affected by seed 
distribution), and disturbance history also affect woody species' seed survival and 
seedling establishment (Leek 1989). Many woody wetland species spread vegetatively, 
or rely on seasonal flooding to distribute their seeds, which may result in concentrations 
along waterways, drift lines, and high water limits, and create a paucity of seeds in 
floodplain areas scoured by seasonal flooding. This concentration of woody and 
herbaceous seeds in areas of the floodplain landscape ultimately contributes to the seed 
bank and vegetation diversity and standing vegetation distribution and structure. In the 
Okefenokee Swamp landscape, surface flow is associated with inflowing northwestern 
streams, the Suwannee and St. Marys River floodplains, and portions of the canoe trails 
that link the prairies and forested regions to these drainages. Berms of peat and live 
vegetation border much of this flow network. Although some of these channels are 
natural topographic lows, many were excavated and have been maintained as boat trails 



750 
during the past 100 years. In many places the vegetation along these trails is a product of 

this maintenance, as peat is elevated, seed banks are exposed, and water and wind 
dispersed seeds are trapped in the berm vegetation. Local and landscape level processes, 
such as fire behavior and water movement during low water periods when peat in 
adjacent areas is exposed, as well as seed and seedling dispersal, are potentially affected 
by this boat trail system. 

Dominant species in the established vegetation may not be well-represented in the 
seed pool for many reasons. In some wetlands, fluctuations in water levels are necessary 
to maintain seed bank and floristic diversity (Leek 1989). Complex relationships among 
the seed bank and established species result where an annual or seasonal drawdown cycle 
occurs. This requires that inundation-tolerant and exposure-tolerant species coexist and 
occur simultaneously in the seed bank. Frequently this concentration occurs along the 
transitional, wetted edge, and not within or outside the wetland (Leek 1989). Many 
species are tolerant of a range of water levels during recruitment (Keddy and Ellis 1985). 
Submerged species germinate almost exclusively under flooded conditions, whereas 
many emergent perennials and mudflat annuals germinate under flooded and drawdown 
conditions (Leek 1989, van der Valk and Welling 1988). Seedling densities are usually 
reduced with prolonged flooding; continuous inundation limits seed survival (Leek 
1989), and reduces seed bank diversity. 

Dispersal mechanisms influence spatial distributions of standing vegetation and 
their propagules. Nearly equal numbers of wind-dispersed (22 species, of which 6 are 
woody) and water-dispersed (26 species, of which 7 are woody) species were present in 



751 
the Okefenokee Swamp seed banks. However, spatial distributions of these species 

differed. Of the wind-dispersed species, 50% were found in all sampled areas, whereas 
only 27% were found in only 1-2 of the sampled areas. Water-dispersed samples were 
more limited in their distributions; 35% were found in all sampled areas, whereas 39% 
were found in only 1-2 of the sampled areas. Wind dispersal increases the likelihood that 
a seed will be distributed away from the parent plant and its seeds, and therefore may 
lessen intraspecific competition upon germination. Although wind-dispersed species 
may be abundantly represented in the seed bank, many of the distributed seeds will fall 
on unsuitable germination sites, and mortality will be high when, if not before, dormancy 
is broken. 

Limitations of hydrochory as the primary seed dispersal mechanism also can 
influence seed survival and therefore community vegetation dynamics. Titus (1990) 
found that distribution of floodplain forest seedlings was correlated with microsite type, 
location, and relationship to floodplain hydrologic environment. Seeds that fall in areas 
with continuous surface water movement will be transported away from the parent plant 
and possibly to suitable germination sites as long as buoyancy is maintained. Seasonal 
inundation results in another seed bank dynamic. Seeds that fall on exposed sediments 
will remain concentrated in place until seasonal flooding removes them. If dormancy is 
broken before dispersal, the seedling must gain sufficient stature to survive inundation 
that will occur with seasonal flooding, and must compete for resources with the parent 
plant and others in the seed rain. Dehiscense of many riverine and floodplain species 
corresponds to low water periods, so that when flooding resumes, seed dormancy has 



752 
broken and the seed is prepared to germinate following water transport (Leek 1989). 

Low water periods in Okefenokee Swamp occur in the late spring and autumn. Seed rain 
of many woody species in the swamp occurs during the late summer and autumn low 
water period. Although seeds falling in the sill-affected area during low water periods 
might successfully germinate on exposed peat, seedling survival depends on achieving 
sufficient height to exceed water levels upon re-flooding. In the impounded area this re- 
flooding occurs with the winter storm fronts and continues into early spring. If 
drawdown to expose these seedlings does not again occur by the growing season, the 
previous year's seedlings will not survive. Greatest seed survival occurs when water- 
dispersed seeds are intercepted by floating debris, concentrate at the edges of receding 
water levels, and settle with drawdown before seed buoyancy declines (Titus 1990, 
Schneider and Sharitz 1988, 1986). Dependence on hydrochory for seed dispersal 
usually limits the spatial extent because of the temporal patterns of hydrologic cycling. 
Therefore, greater incorporation of wind-dispersed seeds into the seed bank over a 
greater spatial and temporal extent is expected, especially if artificially high water levels 
and extended hydroperiods are reduced with sill modification. 

Seed banks may be more diverse where habitat diversity is high (Leek 1989). The 
greater the diversity of site microtopography, the greater the number of species that may 
find suitable germination conditions and subsequently contribute to the standing 
vegetation and seed pool species richness (Titus 1990). Area, treatment, and structural 
zone of origin significantly affected Okefenokee Swamp seed bank diversity in this 
study. Species diversity and richness were greater in exposed conditions, and inter- 



753 
sample variabilities in diversity and richness were higher in submerged conditions. 
Total seedling densities also followed this trend. These trends mirror those of Gerritsen 
and Greening (1989), although densities for individual species differed. Low species 
diversity is not uncommon in submerged seed banks (Leek 1989). 

Seasonal fluctuations in seedling emergence diversity, richness, and total number 
were primarily among sample variances; means were not significantly different. These 
seasonal affects probably reflect the predominant dispersal method, wind, as well as the 
abundance of persistent species in the seed bank. Most of the germinated species 
initially release seeds in the late summer and autumn, and these would have been 
included in the autumn samples. Some species with abundant seed rains are spring seed 
producers, and their seeds are released early in the growing season. By the spring sample 
collection, the seed densities of species producing seeds the previous autumn may have 
declined so that seasonal species differences were apparent, but overall numbers 
remained high because of the additional recent contributions to the pool by spring seed 
producers. Seasonal variability in seed density and diversity may also reflect patchiness 
of the standing vegetation distributions. 

Area was a significant factor in estimating species density, diversity, and 
richness, and differences among areas may also be a function of patchy standing 
vegetation distributions. Overall seed bank species diversity was highest in Chesser 
Prairie; seasonal differences in diversity were most apparent in the sill area, where 
diversity was second to Chesser Prairie in the autumn and lowest of the sampled areas in 
the spring. This seasonal difference may be attributable to the relatively low abundance 



754 
of herbs, most of which are autumn seed producers, and the abundance of shrubs and 
trees, whose seeds are not long-lived in the seed bank after dehiscence. 

Hydrological patterns affect the role seed banks play in wetland vegetation 
dynamics (Leek 1989, van der Valk and Welling 1988, van der Valk and Davis 1979, van 
der Valk and Davis 1978, van der Valk 1981). Presence or absence of standing water is 
the primary "environmental sieve" determining species recruitment or extirpation (van 
der Valk 1981), although the effect of this condition is not uniform for all wetlands (Leek 
1989). Tidal freshwater wetlands experience both inundated and exposed conditions, so 
that the mere presence of water is not necessarily the determining factor in seed survival 
and germination (Leek 1989). In contrast seed banks and hydrologic cycling are vital to 
the long-term survival of marshes of the North American Midwest. Dominant prairie 
wetland species change with water level fluctuations, but all stages of marsh vegetation 
are present in the seed bank, and are renewed with water level fluctuation and dispersal 
from adult plants (Leek 1989, v